United States
             Environmental Protection
             Agency
Oftice ol
Toxic Substances
Washington DC 20460
EPA-560/5-82-006
October, 1982
             Toxic Substances
«>EPA
             Analytical Methods
             for By-Product PCBs-
             Preliminary Validation
             and  Interim Methods
             1000
             1000
             1000
                          ll..
                               '3C12H6CI4


                280 285 290 295 300  305 310  315 320 325

                          Daltons

-------
                                 DISCLAIMER


     This document has been reviewed and approved for publication by the
Office of Toxic Substances, Office of Pesticides and Toxic Substances,  U.S.
Environmental Protection Agency.  Approval does not signify that the contents
necessarily reflect the views and policies of the Environmental Protection
Agency, nor does the mention of trade names or commercial products constitute
endorsement or recommendation for use.

-------
       ANALYTICAL METHODS FOR BY-PRODUCTS PCBs—PRELIMINARY
                  VALIDATION AND INTERIM METHODS
                                By

Mitchell D. Erickson, John S.  Stanley,  Kay Turman,  Gil Radolovich,
     Karin Bauer, Jon Onstot,  Donna Rose, and Margaret Wickham

                    Midwest Research Institute
                    ^  425 Volker Boulevard
                      Kansas City, MO  64110
                              TASK 51

                       INTERIM REPORT NO.  4

                    EPA Contract No. 68-01-5915
                    MRI Project.No. 4901-A(51)

                         October 11, 1982
                                For

               U.S. Environmental Protection Agency
                    Office of Toxic Substances
                       Field Studies Branch
                              TS-798
                      Washington, D.C.  20460

           Attn:  Dr. Frederick W. Kutz, Project Officer
                  Mr. David P. Redford, Task Manager

-------
                                   PREFACE

     This report presents the results of a preliminary method validation ac-
complished on MRI Project No. 4901-A, Task 51, "PCB Analytical Methodology
Task," for the Environmental Protection Agency (EPA Prime Contract No. 68-
01-5915) during the period April 24 to August 31, 1982.

     The document was prepared by Drs. Mitchell D. Erickson (Task Leader) and
John S. Stanley and Ms. Kay Turman, with assistance from Kathy Funk, Cindy
Melenson, and Gloria Sultanik.  The laboratory work was conducted by Kay
Turman, and Donna Rose, with assistance from Steven Turner.  The gas chro-
matography/mass spectrometry analysis was performed by Gil Radolovich,
Margaret Wickham, Jon Onstot, and Arbor Drinkwine.  Statistical analysis of
the data was provided by Karin Bauer.  Editorial comments were provided by
Rudena Mallory and Jeanne Robson.

     The EPA Task Manager, David Redford, has been especially helpful and en-
couraging.  The helpful comments of Ann Carey, Frederick W. Kutz, and John
Smith, all of EPA, are also appreciated.

Z                                                  SEARCH INSTITUTE
                                                     <-
                                                  n  ^
                                             E.  Going,  Head
                                        Environmental Analysis Section
Approved:
James L. Spigarelli, Director
Analytical Chemistry Department
                                    iii

-------
                                  CONTENTS

Preface	iii
Figures	vii
Tables	   ix

     1.   Introduction	    1
     2.   Summary	    2
     3.   Experimental 	    3
               Preparation of PCB stock solutions and working standards.    3
               Gas chromatography/electron impact mass spectrometry. .   .    8
               Determination of PCB response factors (GC/EIMS) 	    8
               Validation of method steps	   19
               Validation with product and product waste samples ....   20
     4.   Method Validation	   24
               Preparation of analytical methods 	   24
               Gas chromatography/mass spectrometry of PCBs	   26
               Validation of selected method steps 	   38
               Validation of product and product waste method with
                 samples	   45
               Discussion	   56
     5.   References	   70
                                            '   /

Appendix A - Supplementary GC/EIMS Data on PCB Congeners 	  A-l

Appendix B - Analytical Method:  The Analysis of Incidentally
  Generated Chlorinated Biphenyls in Commercial Products and Product
  Wastes	B-l

Appendix C - Analytical Method:  The Analysis .of Incidentally
  Generated Chlorinated Biphenyls in Air ... 	  C-l

Appendix D - Analytical Method:  The Analysis of Incidentally
  Generated Chlorinated Biphenyls in Industrial Wastewater 	  D-l

-------
FIGURES
Number
1

2

3

4

5



6

7


8

9

10



11

12

13


Plot of average response factor versus homolog for 77 PCB
congeners 	
Plot of response factor per isomer versus homolog for 77
PCB congeners 	
Plot of response factor per isomer versus homolog for 77
PCB congeners, determined on a single day 	
Retention times of 77 PCB congeners relative to 3,3*4,4'-
tetrachlorobiphenyl-d6 (RRT of 1.00) 	
Capillary gas chromatography/electron impact ionization
mass spectrometry (CGC/EIMS) chromatogram or the cali-
bration standard solution required for quantitation of
PCBs by homolog 	
Reconstructed ion chromatogram for SIM analysis of the
CMA-A sample no. 2110 	
SIM ion plots for monochlorobiphenyls (188 and 190
Daltons) and the 13C6-monochlorobiphenyl surrogate (194
Daltons) in CMA-A sample no. 2110 	
SIM ion plots for dichlorobiphenyls (222 and 224 Daltons)
in CMA-A sample no. 2110 	
SIM ion plots for trichlorobiphenyls (256 and 258 Daltons)
in CMA-A sample no. 2110 	
SIM ion plots for tetrachlorobiphenyls (290 and 292
Daltons), 3,3' ,4,4' -tetrachlorobiphenyl-d6 (298 Daltons) ,
and the 13C12-tetrachlorobiphenyl surrogate (304 Daltons)
in CMA-A sample no. 2110 	
SIM ion plots for pentachlorobiphenyls (326 and 328
Daltons) in CMA-A sample no. 2110 	
SIM ion plots of hexachlorobiphenyls (360 and 362 Daltons)
in CMA-A sample no. 2110 	
SIM ion plots of heptachlorobiphenyls (394 and 396 Daltons)
in CMA-A sample no. 2110 	
Page

27

28

32

36



39

59


60

61

62



63

64

65

66
  Vll

-------
                           FIGURES (continued)

14      SIM ion plots of octachlorobiphenyls (428 and 430 Daltons)
          and the 13Ci2~octachlorobiphenyl surrogate (442 Daltons)
          in CMA-A sample no.  2110	   67

15      SIM ion plots of nonachlorobiphenyl (464 and 466 Daltons)
          in CMA-A sample no.  2110	   68

16      SIM ion plots of decachlorobiphenyl (498 and 500 Daltons)
          and the 13C12-decachlorbiphenyl (510 Daltons) in CMA-A
          sample no.  2110	   69
                                  Vlll

-------
                                   TABLES

Number                                                                Page

   1      Numbering of PCB Congeners	    5

   2      Working Solutions for PCB Response Factors	    6

   3      Approximate Concentration of Individual PCB Congeners in
            Dilute Working Standards	    7

   4      Concentrations of Congeners in PCB Calibration Standards
            (ng/ml) 	    9

   5      Composition of Surrogate Spiking Solution (SS100) Contain-
            ing 13C-Labeled PCBs	   10

   6      Operating Parameters for Capillary Column Gas Chromato-
            graphic System	11

   7      DFTPP Key Ions and Ion Abundance Criteria for Quadrupole
            Calibration	12

   8      Operating Parameters for Quadrupole Mass Spectrometer
            System	13

   9      Operating Parameters for Magnetic Sector Mass Spectrometer
            System	   14

  10      Characteristic Single lion Monitoring (SIM) Ions for PCBs .   15

  11      Limited Mass Scanning (LMS) Ranges for PCBs	16

  12      Characteristic Ions for 13C-Labeled PCB Surrogates	17

  13      Pairings of Analyte, Calibration, and Surrogate Compounds .   18

  14      Commercial Product and Product Waste Stream Samples
            Received for Preliminary Method Validation Studies. ...   21

  15      Preliminary Method Validation Samples 	   22
  16      Comparison of Average Relative Response Factors (RRF) for
            77 Commercially Available PCB Congeners Measured Over
            Several Days as Four Replicates Each Versus Single Mea-
            surements of All Congeners in a Single Day	30
                                     IX

-------
                             TABLES (continued)

Number                                                                Page
  17      Average Relative Response Factors (RRF) for PCB Congeners
            in Solution 1 Measured as Replicates on a Single Day
            and as Single Measurements for Day-to-Day Basis	31
  18      Measured Average Response Factor (RRF) and Corresponding
            Upper and Lower 95% Confidence Limits	34

  19      Relative Response Factors Measured Versus 3,3',4,4'-Tetfa-
            chlorobiphenyl-de by Electron Impact Mass Spectrometry
            Quadrupole (Finnigan 4023) and Magnetic Sector (Varian
            (MAT 311A) Instruments	35

  20      Relative Retention Time (RRT) Ranges of PCB Homologs Versus
            d6-3,3',4,4'-Tetrachlorobiphenyl	37

  21      Recovery Data for Acid Cleanup	40

  22      Recovery Data for Florisil Column Protocol Cleanup	41

  23      Recovery Data for Florisil Slurry Protocol Cleanup	42

  24      Recovery Data for KOH Protocol Cleanup.	43

  25      Recovery Data for Alumina Protocol Cleanup	44

  26      Uncorrected PCB Concentrations ((Jg/g) in CMA-A Samples. .   .  46

  27      Corrected PCB Concentrations (Mg/g) in CMA-A Samples.  ...  47

  28      Uncorrected and Corrected PCB Concentrations  (Mg/g) in
            CMA-E Sample (Dilution Preparation) 	  49

  29      Uncorrected PCB Concentration (Mg/g) in the CMA^A Sample
            Matrix (Internal Standard Calculation)	50

  30      Corrected PCB Concentration (Mg/g) in the CMA-A Sample
            Matrix	51

  31      Uncorrected PCB Concentration (Mg/g) of Spiked CMA-A
            Samples Determined by the Internal Standard Quantitation
            Method	52

  32      Corrected PCB Concentration (Mg/g) of Spiked CMA-A Samples
            Determined by Surrogate Recovery Correction 	  53

  33      PCB Concentration (Mg/g) of CMA-A Samples Heated With Dif-
            ferent Cleanup Procedures (Internal Standard Quantita-
            tion)	54

-------
                             TABLES (continued)

Number                                                                Page

  34      PCB Concentration (|Jg/g) of CMA-A Samples Treated With
            Various Cleanup Procedures (Surrogate Compound Cor-
            receted)	55

  35      Recovery (%) of Carbon-13 Labeled Surrogate Compounds
            From Diarylide Yellow and Phthalocyanine Blue and Green
            Pigments	57
                                     XI

-------
                                  SECTION 1

                                INTRODUCTION

     The Environmental Protection Agency (EPA) is in the process of preparing
rules for regulation of certain polychlorinated biphenyls (PCB)  which are
generated as by-products in the manufacture of commercial products (U.S.  EPA,
1982).  This regulation is under the Toxic Substances Control Act (PL 94-469),
and EPA's Office of Toxic Substances has been assigned the task  of preparing
the rule.

     As part of the rule, EPA is suggesting analytical methods for PCBs in
air (stack gas and fugitive emissions), wastewater, product waste streams,
and final products to assist organizations seeking an exclusion  under this
rule.  To assist EPA in this mission, Midwest Research Institute (MRI) was
asked to prepare appropriate analytical methodologies.  A literature review
and recommendation of general analytical approaches (Erickson and Stanley,
1982; Stanley and Erickson, 1982) constituted the first phase.  The second
phase, reported here, covers initial method validation and preparation of in-
terim methods.  As part of the method validation, four 13C-PCB surrogates were
synthesized and are reported separately (Roth et al., 1982).  The third phase
will involve interlaboratory validation and method refinement.

     This report presents the initial results of method validation for analy-
sis of by-product PCBs in product and product waste samples.  Specifically,
gas chromatography/electron impact mass spectrometry retention time and re-
sponse factor data for 77 PCB congeners for two different gas chromatography/
mass spectrometry systems, recoveries from several proposed cleanup steps,
and recoveries from industrial samples using a variety of the method options
are presented.

-------
                                  SECTION 2

                                   SUMMARY

     The objective of this study was to present EPA with appropriate methodol-
ogies for the analysis of by-product PCBs in commercial products, product
waste streams, wastewaters, and air.  In addition, EPA requested preliminary
analytical studies to provide data in support of the proposed methods.

     This document presents proposed analytical methods for the analysis of
by-product polychlorinated biphenyls in commercial products and product waste
streams (Appendix B), wastewater (Appendix C), and air (Appendix D).  The pro-
posed methods are based on determination of PCBs using gas chromatography/
electron impact mass spectrometry (GC/EIMS).   Capillary column gas chroma-
tography (CGC) and packed column gas chromatography (PGC) are presented as
alternate approaches.  The 13C-labeled PCS surrogates are added to samples
prior to any sample preparation to allow method flexibility for a wide spec-
trum of matrices.  Recovery of the surrogates will allow determination of the
quality of analytical data.  This method is valid only if the surrogates are
thoroughly incorporated into the matrix.

     The analytical method for commercial products and product waste streams
relies heavily on a strong quality assurance program consisting of use of
four 13C-labeled surrogate PCBs, blanks, duplicates, spiked samples, and
quality control samples.  The analytical methods for water and wastewater are
based on EPA Methods 608 and 625, revised to include the use of the 13C-
labeled surrogates.  Likewise, the air method is a revision of a proposed
method for PCBs in air and flue gas emissions.

     This document presents relative response factors (RRF) of 77 PCB congeners
which were used to determine the average RRF for PCBs by homolog.  Statistical
analysis of the data was performed to check the validity of the response
factor data and to extrapolate RRFs for the unavailable congeners.  Relative
retention time (RRT) data for the 77 PCB congeners are also presented.   The
RRF and RRT data were determined on both magnetic sector and quadrupole mass
spectrometer systems.

     Preliminary studies were undertaken to check the validity of the pro-
posed methods for the analysis of PCBs in commercial products and product
waste streams.  Data are presented for analysis, of individual cleanup pro-
cedures as well as for analysis of product and product waste samples.  The
data indicate that the proposed method is applicable and useful for analysis
of the matrices studied.  However, these studies are preliminary and addi-
tional validation is necessary and ongoing.

-------
                                   SECTION 3

                                 EXPERIMENTAL

     The method validation was conducted in three stages:   (a) determination
of GC/EIMS parameters for 77 PCB congeners; (b) validation of individual
method steps with clean matrices; and (c) validation of selected method op-
tions with real samples.

PREPARATION OF PCB STOCK SOLUTIONS AND WORKING STANDARDS

Source of Standards

     Seventy-seven PCB congeners were acquired from Ultra  Scientific, Inc.,
Hope, Rhode Island, and Analabs, North Haven, Connecticut.  Quality control
gas chromatography/flame ionization detection (GC/FID) data for the specific
isomers were requested to verify the 99% purity assigned to these compounds.
The GC/FID data supported the reported purity.  In addition, all available
nuclear magnetic resonance spectra used for specific isomer identification
were requested but not supplied.

Weighing Procedures

     Accurate mass measurement required calibration of a Cahn microbalance
with National Bureau of Standards (NBS) certified masses of 5 and 10 mg.  The
balance was calibrated with the NBS standards followed by calibration of an
in-house working standard mass.  The calibration of the microbalance with the
NBS certified masses was witnessed by a representative of the MRI quality as-
surance office.  The mass of the working standard was measured between all
measurements of individual PCB isomers to ensure that the balance was operat-
ing accurately.  A record of the measured working standard mass was kept in a
laboratory notebook.  The mean value for the working standard was 10.037 ±
0.002 mg (0.02% relative standard deviation).  When all measurements were com-
pleted, the mass of the NBS certified standards was determined as a final mea-
sure of the accuracy of the Cahn microbalance.

Preparation of Solutions

     Preparation of PCB standard stocks began after accurate performance of
the Cahn balance was demonstrated with the certified NBS and daily working
standard.  An aluminum weighing pan was preshaped such that complete transfer
of the weighing pan plus sample could be made directly into the appropriate
dilution vessel.  The Cahn balance was tared to compensate for the weight of
the aluminum boat, and the PCB standards were added via a micro spatula.  The
mass of the particular PCB was determined with the Cahn balance.

-------
     The aluminum pan containing the PCB standard was transferred to the di-
lution vessel using clean forceps, taking care not to spill any of the sample.
The dilution vessel was capped tightly until solvent was added.

     All PCB congeners were dissolved in toluene (Burdick and Jackson, dis-
tilled in glass).  Masses of 0.1 to 5 mg were dissolved in a total of 1.0 ml
toluene while masses of approximately 10 mg and greater were dissolved in
5.0 ml toluene.  The solvent was delivered volumetrically by pipette.  Room
temperature and solvent temperature were recorded at the time of standard
dissolution.  Volumetric pipettes used for solvent delivery were calibrated
so that the most accurate determination of analyte concentration could be
calculated.  Toluene was pipetted into a tared vessel, and the total mass was
measured.  Density of the solvent at the specific room temperature was used
to calculate the actual volume dispensed.  This calibration was performed for
all pipettes used for volumetric delivery of solvent.  The stock solutions
were sonicated in an ultrasonic bath for at least 15 sec after the volumetric
addition of toluene to ensure complete dissolution of the PCBs.  The solution
level was etched on the side of the dilution vessel as a means of detecting
losses by evaporation.

     The individual PCB congeners were referred to by the congener number in-
dicated in Table 1.  The stable labeled PCBs, 3,3',4,4'-tetrachlorobiphenyl-d6,
4-chlorobiphenyl-13C«, 3,3',4,4'-tetrachlorobiphenyl-13C12, 2,2',3,3',5,5' ,6,6,''
octachlorobiphenyl-13C12, and decachlorobiphenyl-^C^ were assigned congener
numbers of 210 to 214, respectively, for the purpose of this work.  Sample
labels were generated in duplicate to identify the specific PCB isomer stock
solution and to document entries in the laboratory notebook.  Table 2 presents
the dilute working solutions that were prepared for determination of the re-
sponse factors for the PCB congeners.  The working solutions were prepared as
10 ml total volume.  Table 3 presents the approximate concentration of each
congener that was in the dilute working standard used for response factor de-
termination.  Tetrachlorobiphenyl-dg was added to 1.0 ml of each solution as
the internal standard.  All stocks were added to the working solutions in vol-
umes of 20, 200, 250, 400, 500, or 1,000 pi.  The syringes were calibrated at
these volumes.  Calibration of the 10-ml volumetric flasks used for working
standards was accomplished by measuring the difference between the mass of
the empty flask and the mass of the flask plus toluene added to the appropri-
ate dilution mark.  The density of toluene at the correct solvent temperature
was used to calculate the final volume of each solution.

     The dilute working solutions were divided into multiple aliquots.  One
hundred micrograms of tetrachlorobiphenyl-d6 was added to each of the 1.0-ml
aliquots of the solutions that were used to establish CGC/EIMS response factors.
The remaining dilute working solutions were stored in at least four crimp seal
vials and refrigerated.  The solvent meniscus was marked in permanent form to
note losses of solvents from evaporation or spills.  All solutions, stock
standards and working solutions, were stored in a refrigerator.  All vials
removed from storage were first brought to room temperature and then sonicated
for at least 15 to 30 sec before removing any of the solution.

-------
TABLE 1. CUMBERING OF PCB CONGENERS3
tJo.
1
2
3
4
5
5
7
3
9
10
11
12
13
14'
15
16
17
13
19
20
22
23
24
25
26
27
23
29
30
31
32
33
34
25
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Structure
HonadilorobiBhCTylj
2
3
4
Q1eh1arBblBh«ny1i
2.2'
2.3
2.3'
2.4
2,4'
2,5
2,6
3.3'
3,4
3.4'
3,5
4.4'
THeMorob
-------
                                 TABLE  2.   WORKING  SOLUTIONS FOR PCB RESPONSE FACTORS

PCB
homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
Total
congeners

Soln.
no. 1
1
11
29
47
121
136
181
195
207
209

10

Soln.
no. 2
2
5
21
44
97
129
171
194
208


9
PCB congener no.
Soln. Soln. Soln. Soln. Soln. Soln. Soln. Soln. Soln. Soln. Soln. Soln.
no. 3 no. 4 no. 5 no. 6 no. 7 no. 8 no. 9 no. 10 no. 11 no. 12 no. 13 no. 14
3
7 8 9 10 4 12 14 15
31 26 24 28 18 33 30
40 49 50 52 53 54 66 61 65 69 72 70,75,77
88 93 101 103 100 104 a 115 87 116 119
128 137 138 141 143 151 139 153 154 155 156
183 185
198 200 202 204
206


97665544333 3

a  Congener no.  112 was added to this  solution but,  on analysis, was  determined to  have  a mass  of 286 and appeared
     to be a diaminotrichlorobiphenyl.   This  congener was  omitted  from any further  consideration.

-------
       TABLE 3.  APPROXIMATE CONCENTRATION OF INDIVIDUAL PCB CONGENERS
      	IN DILUTE WORKING STANDARDS3	

        PCB homolog                             Concentration (pg/ral)
     Monochlorobiphenyl                                   50

     Dichlorobiphenyl                                     50

     Trichlorobiphenyl                                    50

     Tetrachlorobiphenyl                                 100

     Pentachlorobiphenyl                                 100

     Hexachlorobiphenyl                                  100

     Heptachlorobiphenyl                                 100

     Octachlorobiphenyl                                  200

     Nonachlorobiphenyl                                  200

     Decachlorobiphenyl                                  200
a  Tetrachlorobiphenyl-de was added to all solutions as an internal standard
     at ~ 100 M8/ml-

-------
Preparation of Calibration Standard and Spiking Mixtures

     A mixture of 11 congeners was used for calibration.  This solution was
spiked into solvent for protocol step validation experiments and into product
and product waste samples for standard addition experiments.  These congeners
were determined to be the best standards for quantitation calibration based
on the average relative response factor for each PCB homolog, as will be dis-
cussed in Section 5.

     Table 4 presents the composition of the 11-component solutions that are
specified as the calibration standards, CSxxx, where the xxx is used to en-
code the nominal concentration in nanograms per milliliter.   A more con-
centrated solution was diluted as necessary to prepare spiked samples and
the appropriate standards for GC/EIMS analysis.  The internal standard, tetra-
chlorobiphenyl-dg, was added to all standards and final extracts before GC/
EIMS analysis.  The standards contained the four 13C-labeled PCBs that were
added from the spiking solution shown in Table 5.

GAS CHROMATOGRAPHY/ELECTRON IMPACT MASS SPECTROMETRY

     The capillary gas chromatography parameters used are shown in Table 6.
The quadrupole and magnetic sector mass spectrometer parameters used are
shown in Tables 7 through 9.  The characteristic ions for single ion monitor-
ing and limited mass scanning are presented in Tables 10 through 12.

     All data generated for relative response factors and concentration levels
of PCBs in sample extracts were calculated based on the area of the primary
quantitation ion specified in Table 10.  The quantitation ions for the 13C-
labeled monochloro-, tetrachloro-, octachloro-, and decachlorobiphenyl were
194, 304, 442, and 510 Daltons, respectively.  The pairings  of analyte, cali-
bration, and surrogate compounds are presented in Table 13.

DETERMINATION OF PCB RESPONSE FACTORS (GC/EIMS)

     The response factors for 77 PCB isomers were determined by GC/EIMS using
the working standards prepared as described in Tables 2 and  3.  A high reso-
lution capillary column (J&W Scientific Durabond DB-5, 15 m, 0.25 |Jm film
thickness) was used for the separation of the PCB mixtures.   Scanning mass
spectrometry was used to calculate response factors for the  PCB isomers
present in each solution versus a known quantity of tetrachlorobiphenyl-d6.

     The quadrupole GC/EIMS system was tuned daily prior to  any acquisition
of data for PCB response factor calculations.  The system was brought to op-
erating temperature for at least 15 rain.  The fluorocarbon FC-43 was intro-
duced to the ion source, and 176 and 502 Daltons were manually adjusted to a
two-to-one ratio.  This was accomplished by adjusting the multiplier voltage
to 300 mV while monitoring 176 Daltons.  A selected ion monitor acquisition
was set up for 176 and 502 Daltons with a variance of 1 Dalton.  The ratio of
the two values was tuned to the two-to-one ratio as described above.  The mass
spectrometer was operated in the normal full scan acquisition mode after tun-
ing with the FC-43.  Approximately 100 ng of decafluorotriphenylphosphine was
injected and the ratio of the values of 198/442 was monitored.

-------
 TABLE 4.   CONCENTRATIONS OF CONGENERS IN PCB CALIBRATION STANDARDS (ng/ml)a

Homolog
1
1
2
3
4
5
6
7
8
9
10
4
1
4
8
10
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
210 (IS)
211 (RS)
212 (RS)
213 (RS)
214 (RS)
CS1000
1,040
1,000
1,040
1,040
1,520
1,740
1,920
2,600
4,640
5,060
4,240
255
104
257
407
502
CS100
104
100
104
104
152
174
192
260
464
506
424
255
104
257
407
502
CS050
52
50
52
52
76
87
96
130
232
253
212
255
104
257
407
502
CS010
10
10
10
10
15
17
19
26
46
51
42
255
104
257
407
502

a  Concentrations given as examples only.

-------
        TABLE 5.   COMPOSITION OF SURROGATE SPIKING SOLUTION (SS100)
                        CONTAINING 13C-LABELED PCBsa

Congener
no.
211
212
213
214
Concentration
Compound (pg/ml)
(I1, 2', 3
(13C12)3
(13C12)2
1 ,4' ,5',6'-13C6)4-chlorobiphenyl
,3' ,4,4' -tetrachlorobiphenyl
, 2 ' , 3, 3 ' , 5, 5 ' , 6, 6 ' -octachlorobiphenyl
(13C12)decachlorobiphenyl
104
257
395
502

a  Concentrations given as examples only.
                                  10

-------
TABLE 6.  OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM
        Parameter
          Value
Gas chromatograph

Column


Liquid phase

Liquid phase thickness

Carrier gas

Carrier gas velocity

Injector

Injector temperature

Injection volume

Initial column temperature

Column temperature program

Separator

Transfer line temperature
Finnigan 9610

15 m x 0.255 mm ID
Fused silica

DB-5 (J&W)

0.25 |Jm

Helium

45 cm/sec

On-column (J&W)

Optimum performance

i.o Mib

110°C (2 min)C
                       .  d
110° to 325°C at 10°C/min

None

280°C
a  Measured by injection of air or methane at 270°C oven temperature.

b  For on-column injection, follow J&W instructions regarding injection tech-
     nique.

c  With on-column injection, the initial temperature equals the boiling point
     of the solvent; in this instance toluene.

d  C12Cl1o elutes at 270°C.  Programming above this temperature ensures a
     clean column and lower background on subsequent runs.

e  Fused silica columns may be routed directly into the ion source to pre-
     vent separator discrimination and losses.
                                    11

-------
  TABLE 7.  DFTPP KEY IONS AND ION ABUNDANCE
      CRITERIA FOR QUADRUPOLE CALIBRATION

Mass               Ion abundance criteria
197                Less than 1% of mass 198
198                100% relative abundance
199                5-9% of mass 198

275                10-30% of mass 198

365                Greater than 1% of mass 198

441                Present but less than mass 443
442                Greater than 40% of mass 198
443                17-23% of mass 442
                      12

-------
   TABLE 8.  OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER SYSTEM

      Parameter                                           Value


Mass spectrometer                                  Finnigan 4023

Data system                                        Incos 2400

Scan range                                         95-550

Scan time                                          1 sec

Resolution                                         Unit

Ion source temperature                             280°C

Electron energy                                    70 eV

Trap current                                       0.2 mA

Multiplier voltage                                 -1,600 V

Preamplier sensitivity                             10 6 A/V
a  Filaments should be shut off during solvent elution to improve instrument
     stability and prolong filament life, especially if no separator is used.
                                  13

-------
 TABLE 9.  OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER SYSTEM


      Parameter                                           Value



Mass spectrometer                                  Finnigan MAT 311A


Data system                                        Incos 2400


Scan range                                         98-550


Scan mode                                          Exponential


Cycle time                                         1.2 sec


Resolution                                         1,000


Ion source temperature                             280°C

               o
Electron energy                                    70 eV


Emission current                                   1-2 mA


Filament current                                   Optimum


Multiplier                                         -1,600 V



a  Filaments should be shut off during solvent elution to improve instrument
     stability and prolong filament life,  especially if no separator is  used.
                                     14

-------
     TABLE 10.  CHARACTERISTIC SINGLE ION MONITORING (SIM) IONS FOR PCBs

Homolog
Ci2H9Cl
Ci2H8Cl2
C12H7C13
cl2Hecl4
C12H5C15
C12H4C16
C12H3C17
Ci2H2Clg
C12HClg
Cl2CllO

Primary
188 (100)
222 (100)
256 (100)
292 (100)
326 (100)
360 (100)
394 (100)
430 (100)
464 (100)
498 (100)
Ion (relative intensity)
Secondary
190 (33)
224 (66)
258 (99)
290 (76)
328 (66)
362 (82)
396 (98)
432 (66)
466 (76)
500 (87)

Tertiary
a
226 (11)
260 (33)
294 (49)
324 (61)
364 (36)
398 (54)
428 (87)
462 (76)
496 (68)

Source:  Rote JW, Morris WJ.  1973.  Use of isotopic abundance ratios in
         identification of polychlorinated biphenyls by mass spectrometry.
         J Assoc Offie Anal Chem 56(1):188-199.
                                                /
a  None available.
                                      15

-------
            TABLE 11.   LIMITED MASS SCANNING (IMS) RANGES FOR PCBs

Compound
C12H9C11
Ci2HgCl2
C12H7C13
Ci2H6Cl4
C12H5C15
C12H4C16
C12H3C17
Ci2H2Clg
C12HC19
C12CllO
C12D6C14
13C612C6H9C1
13C12H6C14
I
13C12H2C18
13Ci2Clio
r*
Mass range (Daltons)
186-190
220-226
254-260
288-294
322-328
356-364
386-400
426-434
460-468
494-504
294-300
192-196
300-306

438-446
506-516

a  Adapted from Tindall GW, Wininger PE.  1980.  Gas chromatography-mass
     spectrometry method for identifying and determining polychlorinated
     biphenyls.  J Chromatogr 196:109-119.
                                   16

-------
        TABLE 12.  CHARACTERISTIC IONS FOR 13C-LABELED PCB SURROGATES

Compound
13C612C6H9C1
13C12H6C14
13C12H2C18
13Ci2Cl10

Primary
194 (100)
304 (100)
442 (100)
510 (100)
Ion (relative intensity)
Secondary
196 (33)
306 (49)
444 (65)
512 (87)

Tertiary
a
302 (78)
440 (89)
514 (50)

a  None available.
                                   17

-------
                      TABLE 13.  PAIRINGS OF ANALYTE, CALIBRATION, AND SURROGATE COMPOUNDS

Analyte
Congener
no.
1
2,3
4-15
16-39
40-81
82-127
128-169
170-193
194-205
206-208
209
Compound
Calibration standard
Congener
no . Compound
2-C12H9Cl 1
3- and 4-C12H9Cl 3
Ci2HgCl2 7
Ci2H?Cl3
C12H6C14
C12H5C15
C12H4C16
C12H3C17
Ci2H2Cls
Cj^HClg
Ci2dio
30
50
97
143
183
202
207
209
2
4
2,4
2,4,6
2, 2', 4, 6
2, 2', 3', 4, 5
2, 2', 3,4, 5, 6'
2, 2', 3', 4, 4', 5', 6
2,2', 3,3', 5, 5', 6, 6'
2, 2', 3, 3', 4, 4' ,5, 6, 6'
C12Clio
Congener
no.
211
211
211
212
212
212
212
213
213
213
214
Surrogate






Compound
13c6-4
13C6-4
13C6-4
13Ci2-3,3'
13Ci2-3,3'
13C12-3,3'
13Ci2-3,3'
13C12-2,2'
13C12-2,2'
13C12-2,2'
13Ci2Cl10

,4
,4
,4
,4
,3
,3
,3


,4'
,4'
,4'
,4'
,3'
,3'
,3'






,5
,5
,5






,5'
,5'
,5'






,6
,6
,6






,6'
,6'
,6'

00

-------
The response of 198 Daltons was 100% full scale and 442 Daltons was adjusted
from 40 to 45% of the base peak.  These criteria were met daily before data
acquisition for response factor calculations was initiated.

     All working standards were brought to room temperature and sonicated be-
fore injection into the GC/MS system.  Solution No. 1 was analyzed daily as a
means of normalizing response factors calculated from day to day.   This al-
lowed some compensation for differences in sensitivity due to subtle changes
in the mass spectrometer operation from day to day.  Also, a solution of tetra-
chlorobiphenyl-de (internal standard) was analyzed separately.   Four replicates
of each working standard were analyzed to calculate variances of the response
factors.  The solutions were sonicated at least 15 sec prior to removal of
sample for injection.  The syringe and needle were rinsed with 200- to 300-|Jl
of toluene between injections.

     The gas chromatograph was operated at 110°C for 2 min,  and programmed at
10°C/min to 325°C.  One microliter injections were made with a J&W on-column
injection system.  Helium carrier flow was adjusted to 45 cm/sec.

     The peak shape of the eluting PCBs was monitored.  If excessive tailing
was noted, the injection end of the fused silica capillary column was removed
and shortened by at least 10 cm.

     Tables 6, 7, and 8 present the instrument and operating parameters that
were used to measure the response factors for the individual PCB isomers in
the working solutions.  Response factors (RF) were calculated using the area
of the peaks for these ions according to the equation:

                     RTr _  APCB MIS
                        ~  A	M	
                           Ais "PCB

       A
 where JPCB =  Area of the quantitation peak of the specific PCB,
        .IS  =  Mass (in nanograms) of the internal standard injected,
       MIS  =  Area of the quantitation peak of the internal standard, and
        PCB =  Mass (nanograms) of the specific PCB injected.

     All relative response factor data were subjected to Student's t-test at
the 95% confidence level to test for significant differences for day-to-day
and solution-to-solution variances.

VALIDATION OF METHOD STEPS

     A limited number of experiments were completed as preliminary validation
steps for the proposed method presented in Appendices B through D.  The ex-
periment included evaluation of several of the cleanup procedures  using solvent
spiked with the 13C-labeled surrogates and a mixture of PCB congeners repre-
senting each of the possible homologs.  The laboratory cleanup procedures fol-
lowed the protocol steps except where noted.  One hexane solvent blank was
analyzed by each procedure with the samples to monitor interferences and con-
tamination.
                                    19

-------
     All samples were analyzed by CGC/EIMS in the full scan mode using the
Finnigan 4023 system.  Tables 6, 7, and 8 present the instrumental parameters.

VALIDATION WITH PRODUCT AND PRODUCT WASTE SAMPLES

Sources of Samples

     Product waste samples were received from Dow Chemical Company (Kent Hodges)
and Vulcan Materials Company (Thomas Robinson) through the cooperation of the
Chemical Manufacturers Association (Robert Fensterheim).   These samples are
aliquots of the materials used for the Chemical Manufacturers Association
(CMA) round robin study (CMA, 1982).  The CMA and associates supplied samples
of chlorinated benzene waste streams, mixtures of chlorinated benzenes, com-
posite waste streams from a chlorinated aliphatic process and a benzene column
bottom sample.  Table 14 presents an inventory of all the samples received.

     Product samples were received from the Dry Color Manufacturers Associa-
tion (J. Lawrence Robinson and Maria DaRoche).  These samples included diarylide
yellow, phthalocyanine green, and phthalocyanine blue pigments that were used
in the Dry Color Manufacturers Association (DCMA) round robin study of an an-
alytical method, reported by the DCMA (1981).  These samples are also included
in the inventory in Table 14.

     The samples supplied by industry are examples of the samples which will
be analyzed using the method in Appendix B.  However, since no attempt was
made to span the range of products and product wastes, the samples analyzed
do not include all matrices which an analyst could encounter.

Experimental Design

     Table 15 presents an overview of the preliminary method validation sam-
ples.  The samples from Table 14 that were used for these studies included
the chlorinated benzene waste stream, CMA-A; the benzene column bottom sample,
CMA-E; and the yellow, blue, and green pigment samples, DCMA-1, DCMA-4, and
DCMA-8, respectively.  Blind quantitation standards and quality control sam-
ples were prepared by the MRI quality control staff either through spiked ad-
dition or by dilution of particular sample matrices.  Other quality control
procedures included the analysis of duplicate samples and blanks and the
validation of cleanup steps.  Two sets of samples were prepared and run at
separate times.  This first sample set us designated by numbers 10 through
110 and the second sample set is designated by numbers 2001 through 2210Q in
Table 15.

     The sample preparations ranged from addition of the 13C-labeled sur-
rogates followed by dilution and injection, to preparation of pigment samples
via sulfuric acid dissolution and hexane extraction or methylene chloride ex-
traction with Florisil cleanup.
                                    20

-------
       TABLE 14.  COMMERCIAL PRODUCT AND PRODUCT WASTE STREAM SAMPLES
             RECEIVED FOR PRELIMINARY METHOD VALIDATION STUDIES3
Sample no.   Quantity
                Sample description
Sample source
CMA-A


CMA-B


CMA-C



CMA-A

CMA-B

CMA-C


CMA-D


CMA-E
100 ml    Chlorinated benzene waste
            stream

100 ml    Mixture of chlorinated benzenes
            with Aroclor 1254 spike

100 ml    Blind spike of CMA-B with the
            addition of 64 ppm of PCB
            isomers

  5 ml    Chlorinated benzene waste
            stream
  5 ml    Mixture of chlorinated benzenes
            with Aroclor 1254 spike
  5 ml    Blind spike of CMA-B with the
            addition of 64 ppm of PCB
            isomers
  5 ml    Composite waste stream sample
            from a chlorinated aliphatic
            process
  5 ml    Benzene column bottoms sample
Dow Chemical Co.


Dow Chemical Co.


Dow Chemical Co.



Vulcan Materials Co.

Vulcan Materials Co.

Vulcan Materials Co.


Vulcan Materials Co.


Vulcan Materials Co.
DCMA-1
DCMA-4
DCMA-6
DCMA-8
DCMA-9
100 g
100 g
100 g
100 g
100 g
Diarylide yellow pigment
Phthalocyanine green pigment
Phthalocyanine blue pigment
Phthalocyanine blue pigment
Phthalocyanine green pigment
DCMA
DCMA
DCMA
DCMA
DCMA

a  Aliquots of CMA-A, CMA-B, and CMA-C were received from two sources, who
     indicated that they were identical.  MRI has assumed that both aliquots
     are the same.
                                  21

-------
                 TABLE 15.  PRELIMINARY METHOD VALIDATION SAMPLES
Sample
 no.
Description
Preparation
Dilution
 factor
  10       CMA-A
  20A      CMA-A
  20B      CMA-A
  60       Hexane blank
 110       CMA-E
2001       Hexane blank
2005       CMA-A3
2010       CMA-A
2020       CMA-A
2025Q      CMA-A
2030       CMA-A + CS002
2040       CMA-A + CS005
2050       CMA-A + CS010,
2060Q      CMA-A + CSXXX
2070Q      CSxxx
2080       Blank,
2090       CMA-Ab
2100       Blank
2110       CMA-A
2120       Blank
2130       CMA-A
2135       DCMA-13
2140       DCMA-1
2150       DCMA-1
2160       DCMA-1
2170Q      DCMA-1
2175       DCMA-4
2180       DCMA-4
2185       DCMA-4'
2190       DCMA-8
2195       DCMA-8J
2200Q      DCMA-8
2210Q      CSxxx
       + no. 11 (50 ppra)
       + no. 11 (20-80 ppm)
0.1 g/10 ml hexane          1/100
0.1 g/10 ml hexane          1/100
0.1 g/10 ml hexane          1/100
None                        None
None                        None
None                        None
0.1 g/1 ml hexane           1/10
0.1 g/1 ml hexane           1/10
0.1 g/1 ml hexane           1/10
0.5-0.2 g/1 ml hexane     *> 1/10
0.1 g/1 ml hexane           1/10
0.1 g/1 ml hexane           1/10
0.1 g/1 ml hexane           1/10
0.1 g/1 ml hexane           1/10
None                        None
DCMA-A                      1/10
DCMA-A (0.1 g)              1/10
DCMA-B                      1/10
DCMA-B (0.1 g)              1/10
Base                        1/10
Base (0.1 g)                1/10
DCMA-B (1.0 g)              1/100
DCMA-B (1.0 g)              1/100
DCMA-B (1.0 g)              1/100
DCMA-B (1.0 g)              1/200
DCMA-B (1.0 g)              1/200
DCMA-B                      1/100
DCMA-B                      1/100
DCMA-B                      1/100
DCMA-A                      1/50
DCMA-A                      1/50
DCMA-A                      1/50
None                        None
a  No surrogates added to assess any background interferences for these
     compounds.

b  Prepared from aliquot received from Dow Chemical Company; all other CMA-A
     samples prepared from aliquot received from Vulcan Materials Company.
                                    22

-------
     The CMA-A and CMA-E samples were each analyzed after 1/10 or 1/100 dilu-
tion, depending on the operating sensitivity of the mass spectrometer.   The
CMA-A chlorinated benzene waste was the most extensively studied matrix of
the available samples.  Sample preparation included the simple dilution de-
scribed above with and without the addition of the four surrogates.   The sam-
ples prepared without surrogates allowed measurement of the background  that
might interfere with the four surrogate compounds.  Duplicate samples of the
CMA-A were analyzed at the same dilution in two separate experiments.  The
CMA-A matrix was also analyzed by standard addition methods with total  spiked
PCB levels of the 11-compound spiking solution (CS050) at approximately 70,
140, and 270 ng/sample.  The CMA-A matrix was also prepared using the sul-
furic acid and ethanolic KOH procedures discussed in Section 9.3.2 of Ap-
pendix D, Cleanup of the Analytical Method:  The Analysis of By-Product
Chlorinated Biphenyls in Commercial Product and Product Wastes (Appendix B).
Variations of the analytical procedures used by the Dry Color Manufacturers
Association (1981) for the analysis of PCBs in various pigments were also ap-
plied to the CMA-A matrix.  The DCMA procedures included acid dissolution fol-
lowed by hexane extraction from the acid (DCMA Preparation A) and Florisil
treatment of the concentrated sample matrix (DCMA Preparation B).  The  homogen-
ization and centrifugation steps required by the DCMA-B procedure were  not
included for the CMA-A matrix.  All samples except those representing blanks
were spiked with the surrogates at levels of 100 to 500 ng and were mixed
thoroughly before beginning the sample preparation.  The typical CMA-A sample
size was 0.1 g.

     The diarylide yellow (DCMA-1), phthalocyanine green (DCMA-4), and  pthalo-
cyanine blue (DCMA-8) pigments were also studied in these preliminary valida-
tions.  The yellow pigment was prepared according to the recommended DCMA-B
procedure, while the green and blue pigments were analyzed following the DCMA-A
procedure.  The preparation of the pigments followed the DCMA procedures except
that the preparation was scaled to 1 g of the yellow pigment instead of the
recommended 5 g.  Blanks, duplicates, and spiked samples were also analyzed
with the set of DCMA samples.

Sample Analysis

     All extracts were analyzed by capillary column gas chromatography/electron
impact mass spectrometry (CGC/EIMS).  Limited mass scanning (LMS) or selected
ion monitoring (SIM) mass spectrometry methods were used for extract analy-
sis, depending on the level of PCBs in the sample extracts and the complexity
of the matrix.  The parameters for analysis via CGC/LMS and CGC/ MS-SIM are
presented in Tables 6 through 13.
                                    23

-------
                                  SECTION 4

                              METHOD VALIDATION


PREPARATION OF ANALYTICAL METHODS

     Analytical methods were prepared for the analysis of by-product PCBs in:

     *  Commercial products and product wastes (Appendix B).

     "'  Air (Appendix C).

     *  Industrial wastewater (Appendix D).

     The analysis of commercial products and product wastes  was covered in
one method since the diversity of matrices in both categories dictates the
same generalized approach.  Air was defined to include stack gases,  fugitive
emissions, and static (room, other container, or outside) air.

Commercial Products and Product Wastes Method

     The objective was to devise an analytical method suitable for enforce-
ment of the regulation concerning by-product PCBs in commercial products and
product wastes.  A detailed rationale for selection of the techniques used in
the method may be found in a separate report (Erickson and Stanley,  1982).

Sample Workup--
     The general approach taken with sample preparation (collection, preser-
vation, extraction, and cleanup) was to provide a framework within which any
reasonable technique could be used.  This is the only acceptable approach to
a method designed to cover "any" matrix. ,

     The use of 13C-labeled recovery surrogates in conjunction with GC/EIMS
was judged to be the most suitable approach (Erickson and Stanley, 1982;
Stanley and Erickson, 1982; Roth et al., 1982).  Using the recovery surrogates,
any losses of PCBs would be detected and could be corrected for in the calcu-
lation of the PCB concentration.
                                    24

-------
     When surrogates are not fully incorporated into the matrix,  their re-
covery will not be representative of the analyte PCB recoveries and recovery
assessment will not be possible.  It is incumbent upon the analyst to recog-
nize this problem and use good scientific judgment with samples that present
a potential problem.  Nonextractable solid polymers may be an example of a
matrix presenting incorporation problems.

PCB Determination--
     As discussed elsewhere (Erickson and Stanley, 1982; Stanley and Erickson,
1982), GC/EIMS appears to be the only acceptable general technique for deter-
mining PCBs in commercial products and product wastes.  The use of either
capillary or packed column GC is permitted.  While strong arguments are pre-
sented for both techniques (Stanley and Erickson, 1982), the analytical results
should be comparable for both techniques provided proper instrument calibra-
tion and operation, analytical, and quality control procedures are followed
as described in the analytical methods.

Quantitation--
     The analytical objective of these methods is to determine if the sample
contains quantifiable PCBs and, if so, at what concentration.  On the assump-
tion that a general knowledge of the congener distribution is important, re-
porting of the concentration by homolog is proposed in the reporting form.
Since a "total PCB" value is also important for summary and comparative pur-
poses, space for this value is also provided on the reporting form.  Other
reporting formats, including "largest isomer or resolvable peak" or "all peaks
greater than a regulatory value," may be easily accommodated using different
tabulations and reporting procedures.

     The PCB concentrations found may be lower than the actual value due to
nonquantitative recovery during extraction or cleanup.  The measured recov-
eries of the surrogates may be used to derive a corrected concentration.  The
analyst must take care that the surrogates are thoroughly incorporated into
the matrix prior to extraction, as discussed above.  The analyst must also
guard against improper corrections because of errors in surrogate quantita-
tion.  These errors may arise from background interferences.  A more thorough .
discussion of quantitation options is presented in a previous report (Erickson
and Stanley, 1982).

Air Method

     The sample collection, preservation, extraction, and cleanup aspects were
taken from the work of Haile and Baladi (1977).  The determination, using GC/
EIMS, is identical to that in the commercial products and product wastes
method except that recovery surrogates are not used.

Wastewater Method                                  I

     The water method is a direct modification of the commercial products and
product wastes method.  As noted in this method, the cleanup and extraction
procedures for EPA Methods 608 (U.S. EPA, 1979b) and 625 (U.S. EPA, 1979a)
may be used.  It is anticipated that, unless conditions dictate otherwise,
most analysts will choose this option.


                                    25

-------
Quality Control

     Each method includes a strong quality control (QC) section.  Given the
complexity of the matrices and complexity of the analyte (209 compounds), the
need for QC is evident.  The various aspects of the QC section were designed
assuming a reasonably large (10 to 100) batch of samples.   For small batches
of samples, the percentage of effort spent on QC can become sizeable.

Alternate Methods

     The methods presented here are intended to be primary methods capable of
generating the best quality data technologically feasible.   The development
and acceptability of secondary (alternate, equivalent, or screening) methods
is not addressed in this report.

GAS CHROMATOGRAPHY/MASS SPECTROMETRY OF PCBs

     Analysis for PCBs requires the use of selected representative standard
compounds since all 209 congeners are not available.  One of the major dis-
advantages of many instrumental methods for PCB analysis is the large vari-
ance of the instrumental response factors for PCB congeners, both within a
homolog and between homologs.  These large differences in response factors
create problems in selecting representative compounds for quantitation pur-
poses.  The response factors of 77 of the possible 209 PCB congeners measured
by GC/EIMS are presented in Tables 16 and 17.  The data suggests that the EIMS
response factor variance among PCB congeners is small relative to other de-
tectors such as the electron capture detector or negative chemical ionization
mass spectrometry.

Relative Response Factors

Quadrupole Mass Spectrometer--
     The relative response factors (RRF) of the 77 PCB congeners were deter-
mined with the Finnigan 4023 quadrupole mass spectrometer as discussed in the
experimental section.  The RRFs were determined two ways to assess the effects
of instrumental variability.  The replicate RRF determinations are the average
of four replicate analyses for each of the PCB congeners, all determined on a
single day to assess the variability of the measurement.  The single RRF de-
terminations are single values from an experiment in which all 14 solutions
containing all 77 congeners were run on one day to minimize instrumental vari-
ability with time.  The data are presented in Appendix A.  The RRFs vary from
approximately 0.2 for decachlorobiphenyl to 4.1 for 2-chlorobiphenyl.  Fig-
ures 1 and 2 present a visual comparison of average replicate and single RRFs
of PCB congeners determined as replicate measurements and as single measure-
ments .
                                    26

-------
          4.5,—
          4.0
          3.5
        o
        J-l
        o
        
-------
                    3.Or—  1
NJ

00
                    2.5
                    2.0
                  U
                  a)
                  U-i


                  SI1-5
                  C
                  o:
                  a
                  co-
                  ai
                    1.0
                    0.5
                                           _L
2

3

2

1

3

1
                                                                   Quadrupole Mass Spectrometer
        _L
J_
                                           34       5       6       7

                                            Homolog (degree  of chlorination)
                                        10
          Figure 2.   Plot of response factor per isomer versus homolog for 77 PCB congeners,  determined on

            a single day.  Each value is representative of single measurements of each  congener with the

            Finnigan 4023 quadrupole mass spectrometer.  This plot indicates the number of  data points that

            overlap  for  specific isomers.

-------
     Table 16 is a summary of the KRF data, where the replicate and single
measurements are averaged over all measured isomers for a homolog.  The rela-
tive standard deviation (Table 16) for the replicate measurements reflects
the variance of the average RRF for each isomer within a homolog.  The abso-
lute area of the internal standard, Congener No. 210, varied by only 4.4% for
all solutions during the single day experiment, as compared to 9.9% for the 7
days required to complete the replicate analyses.  The relative standard devi-
ations based on the four replicate analyses for each of the PCS congeners,
ranged from 0.4 to 9.1%, indicating the reproducibility of the injection for
each solution.

     The average response factors from replicate determinations and single
measurements were subjected to a Student's t-test to determine if there were
any significant differences in measured response factors.  No significant
difference was found for the average response factor values for any of the
PCS homologs except the heptachlorobiphenyl isomers.  A more detailed presen-
tation of the Student's t-test for these values is presented in Table A-2 of
Appendix A.

     A solution of 3,3',4,4'-tetrachlorobiphenyl-de (Congener No. 210) and
Solution No. 1 (Table 2) were both analyzed daily.  The solution of Isomer
No. 210 was used to tune the quadrupole mass spectrometer to the desired
working conditions.  Solution No. 1 was used to determine fluctuations of re-
sponse factors from day to day due to differences in instrumental operating
parameters.  Table 17 presents the data for single day replicate measurements
and day-to-day determination of the response factors for the PCB congeners in
Solution No. 1.  The relative standard deviations calculated for the single
day measurements are considerably lower than the relative standard deviations
from day-to-day analyses.  This is a reflection of the reproducibility on the
part of the operator as well as of the stability of the quadrapole mass spec-
trometer system on a given day.  The relative standard deviation calculated
for day-to-day analyses is indicative of the variation that might be expected
for routine analysis of PCBs.

     A Student's t-test of the Solution No. 1 data (Table 17) indicated that
there are significant differences in response factors from day to day compared
to single day measurements for PCB Congener Nos. 1, 11, 29, and 207.  A more
detailed presentation of this t-test is presented in Table A-3 of Appendix A.

Magnetic Sector Mass Spectrometer--
     The RRFs for the 77 PCB congeners were also determined with a Varian MAT
311A double focusing magnetic sector mass spectrometer.  The RRF values were
determined by single measurements of all congeners on a single day.  The data
are presented in Appendix A and summarized in Figure 3.

Extrapolation of Response Factor Data to All Congeners--
     Since all 209 PCB congeners were not available for determination of RRFs,
it was necessary to extrapolate the average RRF data to project the range of
response factors that might be encountered.  This extrapolation was based on
the assumption that the number of measured isomers (n) are a representative
sample of the entire set of the possible isomers (N).  Thus it was assumed
that the mean for the measured isomers (n) is an unbiased estimate of the
mean for the possible isomers (N).

                                    29

-------
TABLE 16. AVERAGE RELATIVE RESPONSE FACTORS (RRF) FOR 77 COMMERCIALLY AVAILABLE
PCB CONGENERS MEASURED OVER SEVERAL DAYS AS FOUR REPLICATES EACH AND RRF
FOR SINGLE MEASUREMENTS OF ALL CONGENERS IN A SINGLE DAY

PCB homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
No. of
isomers
3
10
9
16
12
13
4
6
3
1

RRF from
replicate
^ a
measurements
3.331
2.027
1.573
0.950
0.720
0.513
0.361
0.253
0.229
0.213
Relative
standard
deviation (%)
19.3
22.0
21.7
18.4
16.7
15.1
6.6
11.9
14.7
2.8

RRF from
single ,
measurement
2.739
2.048
1.592
0.946
0.725
0.500
0.308
0.224
0.188
0.179
Relative
standard
deviation (%)
9.3
15.7
18.1
20.0
17.6
19.1
8.0
17.3
16.2
-


a  Four replicate measurements of the RRF were made for each isomer.  For example,
     the three monochlorobiphenyl isomers were measured four times each.  Hence,
     the RRF and relative standard deviation (%) were calculated from 12 distinct
     values.

b  A single measurement for each of the 77 PCB congeners was completed in a single
     day.  Hence, the RRF reported is the average of one measured RRF for each
     isomer within a homolog.  For example, the RRF and relative standard deviation
     (%) reported for the monochlorobiphenyls were calculated from three distinct
     values.
                                    30

-------
TABLE 17. AVERAGE RELATIVE RESPONSE FACTORS (RRF) FOR PCB CONGENERS IN
SOLUTION 1 MEASURED AS REPLICATES ON A SINGLE DAY AND AS
SINGLE MEASUREMENTS FOR DAY-TO-DAY BASIS3

Single day measurements
Congener
no.
1
11
29
47
121
136
181
195
207
209

RRF
4.073
3.073
2.195
1.062
0.948
0.689
0.383
0.263
0.237
0.213
Std.
deviation
0.118
0.073
0.048
0.059
0.020
0.016
0.009
0.003
0.008
0.006
Relative std.
deviation (%)
2.905
2.363
2.188
5.591
2.127
2.336
2.379
1.184
3.547
2.837
Day-to-day measurements

RRF
3.544
2.733
2.005
1.032
0.955
0.685
0.377
0.270
0.257
0.223
Std.
deviation
0.452
0.300
0.171
0.061
0.036
0.046
0.028
0.022
0.030
0.023
Relative std.
deviation (%)
12.767
10.977
8.535
5.876
3.747
6.688
7.347
8.304
11.757
10.352

a  See Tables 6 and 8 for CGC/EIMS operating conditions.




b  These values calculated from four replicates.




c  These values calculated from 11 separate analyses.
                                        31

-------
   2.50
   2.00
   1.50
  0)
  CO
  c
  o
  ex
  W
  0)
   1.00
   0.50
                                                        Magnetic Sector Mass Spectrometer
                                      4567

                                     Homolog (degree of chlorination)
10
Figure 3.  Plot of response factor per isomer versus homolog  for  77  PCB congeners, determined on a single

  day.  Each value is representative of single measurements of  each  congener with the Varian Mat 311A

  magnetic sector mass spectrometer.  This plot  indicates  the number of data points that overlap for

  specific isomers.

-------
     Table 18 presents the upper and lower 95% confidence limits for the mea-
sured average RRFs.   The extrapolation was necessary for the dichloro- through
octachlorobiphenyl homologs.  The projected upper and lower limits of the av-
erage RRF ranged from 13% for each PCB homolog for trichlorobiphenyls to ap-
proximately 6.5% for the dichlorobiphenyls.  The projected ranges for the
tetrachloro- to octachlorobiphenyls were between these values.

Comparison of Magnetic Sector and Quadrupole RRF Data--
     The two instruments used operate on entirely different principles, so
the results may represent the range of RRFs to be expected from these com-
pounds on different instruments.  Table 19 presents a summary of the data.
As expected, the RRF trends are much different.  Since quadrupole spectrom-
eters discriminate at the high masses, the RRFs for high homologs (higer
masses) are much lower than corresponding values for the magnetic detector
spectrometer.

     A statistical analysis of the data (Student's t-test presented in Table 4
of Appendix A) confirmed that the average RRFs are significantly different
for many of the homologs.  However, the relative standard deviations for the
average RRF of each homolog are not significantly different.  Thus, the ex-
trapolation from a single calibration isomer to all isomers of a homolog should
have similar precision for the two instrument types.

Relative Retention Times

     Relative retention times (RRT) were also calculated from the data gene-
rated for relative response factor measurements with both the quadrupole and
magnetic sector mass spectrometer instruments.  All RRTs for each PCB congener
were calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6.  Figure 4 is a plot
of the RRT data versus PCB homolog.  All data points for the 77 PCB congeners
measured with the quadrupole mass spectrometer are presented.  This plot also
indicates that the relative retention window for the dichloro- to octachloro-
biphenyl homologs may be larger than that actually measured if more of the
possible congeners were present.

     Table 20 presents the observed range of RRTs for the 77 PCB congeners
and additional congeners, identified only by homolog, in an Aroclor mixture
(1016, 1254, 1260).  These RRTs were established using a 15-m fused silica
DB-5 capillary column.  It must be recognized that the RRT windows on other
columns may be substantially different.  Table 20 also presents a projected
RRT window for PCB anaysis.  The overlap of the retention windows of each
homolog must be considered in establishing an instrumental analysis approach
to quantitation of the specific PCB homologs.  This consideration has been
accounted for in the GC/MS requirements for PCB analysis in Appendices B to
D.  The relative retention times of the 77 PCB congeners as determined with
both the quadrupole and magnetic sector mass spectrometers are presented in
tabular form in Appendix A.
                                    33

-------
TABLE 18. MEASURED AVERAGE RELATIVE RESPONSE FACTOR (RRF) AND
CORRESPONDING UPPER AND LOWER 95% CONFIDENCE LIMITS

PCB homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
No. of
possible
isomers
(N)
3
12
24
42
46
42
24
12
3
1
No. of
available
isomers
(n)
3
10
9
16
12
13
4
6
3
1
Average
measured
response
RRF
3.331
2.027
1.573
0.950
0.720
0.513
0.361
0.253
0.229
0.213
Sample std. ,
deviation Lower Upper
(S) limit limit
0.643
0.447 1.896 2.158
0.341 1.366 1.780
0.175 0.877 1.023
0.120 0.654 0.786
0.078 0.474 0.552
0.024 0.326 0.396
0.030 0.231 0.275
0.034
- - -

4-_ / _\ 1.
a  Lower 95% limit = RRF - — /1 - £
b  Upper 95% limit = RRF + —  1 - £
                            V~ I     rl
                            n V
                                      34

-------
 TABLE 19.   RELATIVE RESPONSE FACTORS MEASURED VERSUS 3,3',4,4'-TETRACHLORO
    BIPHENYL-d6 BY ELECTRON IMPACT MASS SPECTROMETRY QUADRUPOLE (FINNIGAN
           4023) AND MAGNETIC SECTOR (VARIAN MAT 311A) INSTRUMENTS

PCB homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
No. of
isomers
measured
3
10
9
16
12
13
4
6
3
1


RRF
Quadrupole_
Mean
2.739
2.048
1.592
0.946
0.725
0.500
0.308
0.224
0.188
0.179
RSD" (%)
9.3
15.7
18.1
20.0
17.6
19.1
8.0
17.3
16.2' -
-

Magnetic
Mean
2.329
1.663
1.167
0.902
0.780
0.640
0.497
0.463
0.467
0.586

sector
RSDa (%)
8.5
13.8
21.3
14.0
17.4
19.4
12.1
15.3
22.5
-

a  Relative standard deviation.
                                  35

-------
CO
           C12H,CI9
          C]2H2CI8
           C,2H3CI7
        S C12H4d6
o

o
X

pq
O
PM
           C12H5CI5
          Cl2H6CI4
           C,2H7CI3
           C,2H8CI2
           C,2H9CI,
                       Relotive Retention Times of PCB Congeners by Homolog

                            Versus 3,3',4,4' Tetrachlorobiphenyl-d0
                                                                                               208*0'    **
                                                                                     202200    lOfl   195 194
                                                                                   185 171
                                                                                  IB!) 181
                                                               IS4 139 153 138 129
                                                         155    136 151  143  141 137  128 156
                                                   	»     «««»«»»««i4  •-
    .„. 121   9/IIS
 104 103ICO 9388 "" 119  t15
-•  •»  «•	• •«•--
                                                75
                                           «,,   6949
                                         54 5° 53  5247 .
                                         • » «	••—
      70
     '61 66     65 77
     »•«     • «-
                                                 2631 33
                                          30  18 24 29  28 21
                                 10   9  8     ,
                                 4   7  5  14 1l"l5
                             23
                                              J_
                                              _L
      _L
_L
0.40     0.50     0.60     0.70     0.80     0.90    1.00

                          Relative retention  time
                                                                              1.10
                                                                              1.20
                                      1.30
                                1.40
      Figure 4.  Retention  times of  77 PCB  congeners relative to  3,3',4,4'-tetrachlorobiphenyl-d5  (RRT of  1.00)

        The  dashed  line indicates  that not  all of  the possible isomers  of a particular  homolog were  measured.

        Relative retention  times were determined on a J&W  DB-5, 15-m  fused silica column in  a Finnigan 4023

        GC/EIMS system.  Temperature program:   110°C for 2 minr then  10°C/min  to 325°C.

-------
       TABLE 20.   RELATIVE RETENTION TIME (RRT) RANGES OF PCB HOMOLOGS
                   VERSUS d6-3.3',4,4'-TETRACHLQROBIPHENYL

PCB
homolog
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
No. of
isomers
measured
3
10
9
16
12
13
4
6
3
1
Observed range
of RRTsa
0.40-0.50
0.52-0.69
0.62-0.79
0.72-1.01
0.82-1.08
0.93-1.20
1.09-1.31
1.19-1.36
1.31-1.42
1.44-1.45
Calibration
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
solution
Observed
RRT3
0.43
0.50
0.58
0.65
0.75
0.98
1.05
1.15
1.19
1.33
1.44
Projected
range of
RRTs
0.35-0.55
0.35-0.80
0.35-1.10
0.55-1.05
0.80-1.10
0.90-1.25
1.05-1.35
1.10-1.50
1.25-1.50
1.35-1.50

a  The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were
     measured versus de-3,3',4,4'-tetrachlorobiphenyl (internal standard)
     using a 15-m J&W DB-5 fused silica column with a temperature program of
     110°C for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec,
     and an on-column injector.  A Finnigan 4023 Incos quadrupole mass spec-
     trometer operating with a scan range of 95-550 Daltons was used to de-
     tect each PCB congener.

b  The projected relative retention windows account for overlap of eluting
     homologs and take into consideration differences in operating systems
     and lack of all possible 209 PCB congeners.
                                     37

-------
Selection of Congeners for a Calibration Standard

     The data generated from the RRF and RRT measurements were used to select
the PCB congeners for an analytical quantitation/calibration standard for
GC/EIMS analysis of PCBs.  Selection of the standard compounds was based pri-
marily on the ratio of the measured response factor to the average response
factor for a particular homolog.  The PCBs with RRFs closest to the average
values were selected as standard compounds.  In addition, the RRT was con-
sidered to assure that the selected PCB congeners did not coelute.  Two mono-
chlorobiphenyls were selected for the calibration standard because the aver-
age RRF and RRT did not clearly coincide with any of the three possible
isomers.  One isomer (2-chlorobiphenyl) had a substantially different RRF.
This isomer was quantitated separately.  4-Chlorobiphenyl was selected as the
calibration isomer for the two remaining isomers.  Figure 5 is a CGC/EIMS
chromatogram of the 11-component PCB calibration standard.  The composition
of this solution is identified in Tables 4 and 20 along with the observed RRT
of each of the 11 congeners.

VALIDATION OF SELECTED CLEANUP STEPS

     As part of the overall method validation, several of the cleanup tech-
niques were validated.  A mixture of the 11 calibration standard congeners
and three recovery surrogates (the 13C-octachlorobiphenyl was unavailable for
these experiments) was diluted in an appropriate solvent and then subjected
to the cleanup procedures as described in Appendix B.  After the cleanup, the
internal standard was added and the volume adjusted.  The samples were analyzed
by CGC/EIMS using a quadrupole spectrometer operated under the condition listed
in Tables 6 through 8.  Data were collected in the full scan mode and quanti-
tated using the primary ions listed in Table 10 and the congener pairs listed
in Table 13.  A blank was run through the procedure alongside the recovery
spikes.  As expected, no PCBs except the internal standard were observed in
the blanks.

     The results for the 11 calibration congeners were calculated as percent-
age recovery.  Tables 21 through 25 present the uncorrected recoveries, cal-
culated using Equation 12-1 of Appendix B, using the internal standard (Con-
gener No. 210); the actual percentage recoveries of the 13C-labeled recovery
surrogates, calculated using Equation 12-2 of Appendix B; and the corrected
recoveries of the calibration congeners, calculated using Equation 12-3 of
Appendix B.

     Inspection of Tables 21 to 25 reveals that the accuracy of the corrected
recoveries is higher than for the uncorrected recoveries (104% versus 77%
average).  On the other hand, the precision of the uncorrected recoveries is
slightly higher than for the corrected recoveries (Il7o versus 9% relative stan-
dard deviation average).  This is the expected trend since the uncorrected
recovery relies on two GC/MS measurements  (area of the PCB congener peak and
area of the internal standard peak) and the corrected recovery relies on those
two values and the area of the surrogate peak.  Thus, these results indicate
that accuracy is improved by recovery correction, at a sacrifice of precision.
                                    38

-------
          100.0-1
           me
u>






















rH
r*1*
c
CU
XI
CX
•H
XI
0
M
0
rH
XI
o
O
C
•8
CN
6I8
|
523 575 H

GK3
10:0:)
VO
rH
1
C C
O> 01
x; x; -
ex ex
XI XI ^
O O 0)
M M -U
o o cd
rH rH 60
XI XI O
CJ O l-i
0 0 rl
C C 3
O O CO
S S ^
i I
•* 
c




















rH

C
01
-. XI
CX
0) -rl













rH

C
0)
XI
CX
XI
0
o
rH
XI
O
cd

XI XI 4J
C
•r-
i. O
J-l
XI 0
c
1-
c
1—
rH
XI
a
•H
XI rl
0)
H

vO

v^-
•.
CN
OH
*f
C
i
vO
CN CN
j
I.I



0 798
«
..,L82L
r
C'JO
13:20
CN

07'1

.


                                                 936
rH Cd
cd oo
C 0
O> t-l
4-1 3
C to

v~' u
vO CO
'^ rH
1 1
rH rH
r^"> r**»
C C
0) 0)
XI XI
ex ex
•H -rH
f O ^Q
0 0
)->
•U C
OJ OJ
•H rC
1 CX
•H
*3" Xi
- O
^J- 1-1
- o
- rH
CO XI
« CJ
ro cd
^
o>
SB
1

vO
"
"1

XI
CX
•rl

O

0
rH

O
cd

CX
cu

1
vO
M
»•
m
VI
-
"*,
^t
f^>






1101


,__(
^
rH C
>i O>
c x:
01 CX
XI -H rH
ex xt >->
•rl 0 C
Xi M CiJ
0 0 XI
1-1 rH CX
O X! -H
rH 0 XI
xi cd o
a C rl
cd o o
4J £7^ rH
O 1 X!
0 CJ
I vo cd
r, O
vO vO 0)
" « Q
vo m

•. »
in ^d"
^ n
in ^
•v n
•. to
CO CO
v> n
CO CO
«i *>
f •.
CN CN
CN CN


^ 1208 I3IG
CN

1178
Jfl





CN

0
CO
,__j
1
rH

'C
0>
XI
ex
•H
x>
0

0
rH
XI
o
cd
a
o>
o









































,— s
cu
4J
cd
60
o
l-i
1-1
3
to
N-X























i ' l
2f>:&) 23:20
                                                                                               112592.
                                                                                                SCA:I
                                                                                                TIlli:
     Figure  5.  Capillary  gas  chromatography/electron impact ionization mass  spectrometry (CGC/EIMS)
       chromatogram or  the calibration standard solution required for quantitation  of  PCBs  by homolog,
       This  chromatogram includes PCBs representative of each homolog, three  ^-^c-labeled  surrogates,
       .and the deliberated  internal .standard.   The concentration of all components and  the CGC/EIMS
       parameters  are presented in Tables 4,  5, 6 and 9.

-------
                                TABLE 21.   RECOVERY DATA FOR ACID CLEANUP'

Congener no .
1
3
7
30
50
97
143
183
202
207
209
X
Standard deviation
Relative standard deviation (%)
211
212
214
X
Standard deviation
Relative standard deviation (%)
Total spike
PCB homolog level (|Jg)
Monochlorobiphenyl
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nona chlo rob iphenyl
Decachlorobiphenyl



13Ce-nionochlorobiphenyl
13Ci2~tetrachlorobiphenyl
13C 12 -decachlorob iphenyl



0.52
0.50
0.52
0.52
0.76
0.87
0.96
1.30
2.30
2.50
2.10



2.60
5.30
10.20



Spike 2 (%
Uncorrected
100.0
83.4
82.5
NQ
78.0
99.5
81.2
85.9
80.7
83.2
87.3
86.2
7.5
9
70.2
87.1
91.3
82.9
11.2
13
recovery)
Corrected
142.4
118.8
117.5
NQ
89.6
114.2
93.2
98.5
88.4
91.1
95.7
104.9
17.7
17
-
-
-

Blank
NDC
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

-
-
ND
ND
ND
-
-


a  Spike No. 1 not analyzed.




b  Corrected via surrogate response.




c  Not detected.




d  Large background signal prevented  quantitation of the compound.




e  Not applicable.

-------
TABLE 22. RECOVERY DATA FOR FLORISIL COLUMN CLEANUP


Congener no.
1
3
7
30
50
97
143
183
202
207
209
X
Standard deviation
.P, Relative standard
M deviation (%)
211
212
214
X
Standard deviation
Relative standard
deviation (%)

PCB homolog
Monochlorobiphenyl
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl




13C6-monochlorobiphenyl
13C12-tetrachlorobiphenyl
13C12-decachlorobiphenyl




Total spike
level (fjg)
0.52
0.50
0.52
0.52
0.76
0.87
0.96
1.30
2.30
2.50
2.10




2.60
5.30
10.20




Spike 1 (%
Uncorrected
57.9
63.0
66.0
69.4
70.7
73.4
72.6
76.6
77.8
78.1
77.7
71.2
6.7
9

63.9
43.2
75.9
61.0
16.5
27

recovery)
Corrected
90.6
98.6
103.2
160.5
163.6
169.7
168.1
117.2
102.5
102.9
102.4
130.9
35.8
27

-
-
-
-

Spike 2 (%
Uncorrected
54.9
58.3
60.0
62.3
62.4
66.1
67.0
72.3
72.3
70.5
72.8
65.4
6.2
10

57.6
47.9
69.6
58.4
10.9
19

recovery)
Corrected
95.4
101.2
104.4
130.0
130.3
138.1
140.1
151.0
103.8
101.3
104.5
118.2
19.8
17

-
-
-
-


Blank
NDb
ND
ND
ND
ND
ND
ND
ND
ND
ND
NDc
-
-
-

ND
ND
ND
-
-
-


a Corrected via surrogate response.
b Not detected.
c Not applicable.















-------
                                  TABLE 23.   RECOVERY DATA FOR FLORISIL SLURRY CLEANUP

Congener no.
1
3
7
30
50
97
143
183
202
207
209
X
Standard deviation
Relative standard
deviation (%)
211
212
214
X
Standard deviation
Relative standard
deviation (%)
PCB homolog
Mpnochlorobiphenyl
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pent a chlo r ob ipheny 1
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nona chlorob ipheny 1
Decachlorobiphenyl




13C6-monochlorobiphenyl
13C12-tetrachlorobiphenyl
13C12-decachlorobiphenyl




Total spike
level ((jg)
0.52
0.50
0.52
0.52
0.76
0.87
0.96
1.30
2.30
2.50
2.10




2.60
5.30
10.20




Spike 1 (%
Uncorrected
80.5
81.2
87.5
NQC
90.0
96.0
95.5
95.1
97.2
95.1
96.2
91.4
6.3
7

83.9
98.3
106.5
92.5
7.5
8

recovery)
Corrected"
96.0
96.8
104.4
NQ
91.6
97.6
97.2
96.8
91.2
89.4
90.4
95.1
4.5
5

_
-
-
-
-
-

Spike 2 (%
Uncorrected
71.1
72.7
75.0
76.4
80.1
83.5
82.0
79.8
88.8
87.6
83.7
80.1
5.8
7

76.7
89.4
87.9
84.7
6.9
8

recovery)
Corrected
92.9
94.8
98.1
85.5
89.6
93.5
91.6
89.3
101.0
99.5
95.2
93.7
4.7
5

_
-
-
-
-
-

Blank
NDb
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND,
Q

-
-

ND
ND
ND
-
-
-


a  Corrected via surrogate response.




b  Not detected.




c  Large background signal prevented quantitation of this compound.




d  Not applicable.

-------
                                            TABLE 24.  RECOVERY DATA FOR KOH CLEANUP
CO

Congener no.
1
3
7
30
50
97
143
183
202
207
209
X
Standard deviation
Relative standard
deviation (%)
211
212
214
X
Standard deviation
Relative standard
deviation (%)
PCB homolog
Monochlorobiphenyl
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl




13Ce-nionochlorobiphenyl
13C12-tetrachlorobiphenyl
13C!2-decachlorobiphenyl




Total spike
level (pg)
0.52
0.50
0.52
0.52
0.76
0.87
0.96
1.30
2.30
2.50
2.10




2.60
5.30
10.20




Spike 1 (%
Uncorrected
60.2
69.0
73.5
75.0
79.7
85.8
84.0
81.2
89.2
88.2
69.9
77.8
9.1
12

72.9
89.9
78.9
80.6
8.6
11

recovery)
Corrected"
82.6
94.6
100.8
83.5
88.7
95.4
93.4
90.3
113.0
111.8
88.6
94.8
10.2
11

_
-
-
-
-
-

Spike 2 (%
Uncorrected
67.7
73.6
77.5
77.6
80.7
85.0
85.0
81.3
89.3
87.6
71.9
79.7
6.8
9

75.1
87.0
76.0
79.4
6.6
8

recovery)
Corrected
90.1
98.0
103.3
89.2
92.9
97.7
97.7
93.5
117.5
115.3
94.6
99.1
9.5
10

_
-
-
-
-
-

Blank
NDb
ND
ND
ND
ND
ND
ND
ND
ND
ND
NDc
-
-
-

ND
ND
ND
-
-
-


    a  Corrected via surrogate response.




    b  Not detected.




    c  Not applicable.

-------
TABLE 25. RECOVERY
DATA FOR ALUMINA
CLEANUP




Total spike Spike 1 (%
Congener no.
1
3
7
30
50
97
143
183
202
207
209
X
Standard deviation
-p~ Relative standard
deviation (%)
211
212
214
X
Standard deviation
Relative standard
deviation (%)
PCB homolog level (pg) Uncorrected
Monochlorobiphenyl
Monochlorobipuenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl




13C6-monochlorobiphenyl
13Ci2~tetrachlorobiphenyl
13Ci2~decachlorobiphenyl




0.52
0.50
0.52
0.52
0.76
0.87
0.96
1.30
2.30
2.50
2.10




2.60
5.30
10.20




63.1
60.0
67.9
NQC
67.2
70.4
69.4
75.8
76.8
77.3
74.0
70.2
5.9
8

64.8
69.1
83.2
72.4
9.6
13

recovery)
Corrected"
97.1
92.2
104.8
NQ
97.2
101.9
100.4
109.7
92.2
92.9
88.9
97.8
6.5
7

-
-
-
-
-

Spike 2 (%
Uncorrected
61.1
58.4
66.4
NQ
66.3
68.3
67.5
75.1
75.3
76.8
78.3
70.1
6.9
10

60.7
64.9
84.2
69.9
12.5
18

recovery)
Corrected
101.0
96.2
109.4
NQ
102.2
105.4
104.2
115.8
89.5
91.2
93.0
100.8
8.4
8

-
-
-
-
-


Blank
NDb
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
U
-
-

ND
ND
ND
-
-
-


a Corrected via surrogate response.
b Not detected.
c Large background
d Not applicable.

signal prevented quantitation


of this


compound.














-------
     The preliminary data presented here contain an apparent anomaly:   the
low recovery of the 13C-tetrachlorobiphenyl surrogate (Congener No.  212) from
the Florisil column cleanup.  These two data points contribute substantially
to the imprecision of the surrogate recoveries and induce some very high (130
to 177%) corrected recoveries for the tri- through hepta- compounds.  The ex-
periment should be repeated.

VALIDATION OF THE PRODUCT AND PRODUCT WASTE METHOD WITH INDUSTRIAL SAMPLES

Strategy

     Selected samples, obtained from industrial sources, were subjected to a
variety of sample preparations as listed in Table 15 and then analyzed by
CGC/EIMS.  This section presents the results of this preliminary validation
and, where possible, compares our values with those of previous analyses of
the same sample.  The results for quality control samples are also reported.

     The most extensively studied matrix was the CMA-A chlorinated benzene
waste stream sample.  This particular sample was chosen because of the wide
distribution of PCB homologs (mono- through decachlorobiphenyls).   Sample
preparation with this matrix included simple dilution, treatment with sul-
furic acid, Florisil, and saponification with ethanolic potassium hydroxide.
The CMA-A samples were analyzed in duplicate in two sets of experiments.  The
11 PCB congeners used for calibration purposes were spiked into the CMA-A
matrix for standard addition experiments.  Blind spiked samples and quantita-
tion standards, prepared by the MRI quality control personnel as analytical
performance checks, were analyzed along with the other samples.

First Sample Set

     Tables 26 and 27 present the uncorrected and corrected concentrations
found for CMA-A samples in preliminary studies of the application of the pro-
posed methods for commercial products and product wastes.  Sample 10 was
analyzed without surrogates to approximate the analytical procedure used by
most other laboratories.  As anticipated, the uncorrected values compare well
with 20A and 20B, while the corrected values are slightly lower than the
values for 10.  Both corrected and uncorrected values for the duplicate sam-
ples 20A and 20B are in agreement.  The values for samples 10, 20A, and 20B
average about 400 (Jg/g.  These values are higher than the mean of 280 |Jg/g
reported in the CMA round robin but are in good agreement with the values
(402 |Jg/g) reported by the sample supplier (Appendix E of Pittaway and Horner,
1982).  The homolog distribution of our data agrees in general with the
CMA data and the data that accompanied the samples.

     Sample 110 (CMA-E) was determined to contain about 18 pg/g PCB (Table
28) mostly as the dichloro homolog.  These results are slightly higher than
the CMA round robin data, which had a mean reported value of 9 pg/g.  The
isomer distribution agrees with most of the CMA round robin data (Pittaway
and Horner, 1982).
                                    45

-------
       TABLE 26.  UNCORRECTED PCS CONCENTRATIONS (pg/g) IN CMA-A SAMPLES
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
Total
211
212
214
PCB
homolog
1
1
2
3
4
5
6
7
8
9
10

1
4
10
10
Dilution,
no surrog.
9
19
64
55
60
50
56
60
0
0.9
9.3
414
NSa
NS
NS
20A
Dilution
11
21
70
52
63
40
48
84
0
0
20
408
96b
108
154
20B
Dilution
10
19
64
49
55
36
38
68
0
0
20
358
94
97
152

a  No surrogates added.




b  Surrogate recovery (percent).
                                    46

-------
            TABLE 27.  CORRECTED PCS CONCENTRATIONS (Mg/g) IN CMA-A SAMPLES

Congener
no.
1
3
7
30
50
97
143
183
202
207
209
Total
PCB
homo log
1
1
2
3
4
5
6
7
8
9
10

10
Dilution,
no surrog.
NSa
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS

20A
Dilution
11
22
73
49
58
37
44
78
0
0
13
385
20B
Dilution
11
21
68
50
57
37
39
70
0
0
13
366

a  No surrogates added.
                                   47

-------
Second Sample Set

CMA Product Waste Samples--
     The corrected and uncorrected concentrations of the PCB homologs for
duplicate CMA-A samples from a more extensive study are presented in Tables
29 and 30.  Sample 2005 was spiked only with the internal standard so that
any interferences corresponding to the 13C-labeled PCBs could be measured.
Samples 2010 and 2020 are duplicate samples of CMA-A.  The four surrogate
compounds were added to approximately 0.1 g of each sample.   The mixture was
diluted to 1.0 ml and the internal standard added.  Sample 2025Q is a sample
that was submitted for PCB analysis by the MRI quality control department.
This sample was weighed by QC personnel and the final preparation completed
as described for the previous samples.  The MRI QC coordinator calculated the
final concentration for 2025Q from the extract concentration of each PCB
homolog and weight of the CMA-A sample recorded in the QC laboratory record
book.  The surrogate-corrected values reported for samples 2010 through 2025Q
are in good agreement with the total PCB concentration and homolog distribu-
tion reported in the CMA round robin (Pittaway and Horner, 1982).

     Tables 31 and 32 present the data from a standard addition experiment
with the CMA-A sample matrix.  The 11 PCB congener calibration standard was
added to three separate aliquots of the CMA-A matrix to give spike levels
ranging from approximately 20 to 100 |jg of the monochlorobiphenyl and 50 to
200 (Jg of decachlorobiphenyl.  Samples 2030, 2040, and 2050 were prepared in
the analytical laboratory.  Sample 2060Q was prepared as a blind spike of the
CMA-A matrix by MRI quality control personnel.  The uncorrected amount found
did not increase linearly with the spike level.  In fact, at the highest spike
level (Sample 2050) the amounts found for each homolog were less than the
spike.  No explanation is immediately available for this data trend, although
the low recoveries of the 13C-octa- and tetrachlorobiphenyl surrogates indi-
cated that the data are at best marginally valid.

     Tables 33 and 34 present data for CMA-A samples that were subjected to
three different cleanup methods (concentrated H2S04, Florisil column chro-
matography, and saponification with alcoholic KOH).  The data from the sul-
furic acid cleanup was difficult to interpret because of interferences.  As
noted previously (Erickson and Stanley, 1982), the acid cleanup results in
large losses of lower chlorinated PCB homologs.  The poor recoveries of the
surrogates shown in Table 33 are clearly outside of the QC criteria in Sec-
tion 14.2.2 of Appendix B and indicate that the analyses are invalid.  These
results would not be reported as analyses for compliance with the proposed
regulation.

     All of the blank samples (2001, 1080, 2100, and 2120) were analyzed
along with the sample discussed above and found to contain no detectable
PCBs.
                                    48

-------
            TABLE 28.  UNCORRECTED AND CORRECTED PCB CONCENTRATIONS (pg/g)
                        IN CMA-E SAMPLE (DILUTION PREPARATION)
Congener
  no.
  PCB
homolog
    110
Uncorrected
   110
Corrected
   1
   3
   7
  30
  50
  97
 143
 183
 202
 207
 209

Total

 211
 212
 214
   1
   1
   2
   3
   4
   5
   6
   7
   8
   9
  10
   1
   4
  10
    1.2
    1.8
   10.5
    0
    0
    0
    0.02
    0
    0.05
    0
    0.06

   13.4
  103/91
  151
   1.5
   2.4
  13.8
   0
   0
   0
   0.02
   0
   0.03
   0
   0.04

  17.7
     b
a  Surrogate recovery (percent).

b  Not applicable.

c  Samples run twice on magnetic sector instrument for low and high masses.
     Congener no. 212 monitored in both runs.
                                     49

-------
     TABLE 29.  UNCORRECTED PCB CONCENTRATION ((Jg/g) IN THE CMA-A
             SAMPLE MATRIX (INTERNAL STANDARD CALCULATION)

PCB
homolog
CGC/EIMS analysis date
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB
Recovery
13C6-monochlorobiphenyl
13C12-tetrachlorobiphenylc
13C12-octachlorobiphenyl
13C12~decachlorobiphenyl
CMA-A CMA-A CMA-A
2005 2010 2020
8/4/82 8/4/82 8/5/82
26
35
17
20
32
29
18
5.4
2.6
12
197
(%) of Surrogate
NSa
NS
NS
NS
23
28
14
31
29
23
12
4.1
2.2
10
176
Compounds
64
96
73
68
37
41
46
33
29
21
12
3.4
2.0
9.7
234

84
96
67
69
CMA-A
2025
8/5/82
40
48
50
36
31
22
14
4.2
3.5
11
260

89
101
72
73

a  NS = no surrogate added.

b  Final concentration determined from sample weight recorded by QC
     coordinator.

c  302 Daltons used for quantitation.
                              50

-------
                                                        J8KCJOHS g'ffi'CAL L!^
                                                         801 L ST., S.W., TS-7^.;
                                                         KWSKINGTON,  D.C. 2Q«Q
 TABLE 30.   CORRECTED PCB CONCENTRATION ((Jg/g) IN THE
                   CMA-A SAMPLE MATRIX

PCB
homolog
CGC/EIMS analysis date
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB
CMA-A
2010
8/4/82
37
44
15
33
30
24
16
5.4
3.1
15
223
CMA-A
2020
8/5/82
44
48
47
34
30
21
18
4.9
3.0
14
264
CMA-A,
2025QD
8/5/82
44
53
49
34
31
22
19
5.7
4.8
16
280

a  NS = no  surrogates added.

b  Final  concentration determined  from sample weight
     recorded by QC coordinator.
                           51

-------
                                      TABLE 31.  UNCORRECTED PCB CONCENTRATION (pg/g) OF SPIKED CHA-A SAMPLES DETERMINED RY THE
                                                                INTERNAL STANDARD QUANTITATION METHOD
Ln
	 ,
CMA-A 2030
Total sample
PCB homolog concentration
CGC/EIMS analysis date
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB

13Cg-monochlorobiphenyl
13Ci2-tetrachlorobiphenyl
13C 12 -Octachlorobiphenyl
13Ci2-decachlorobiphenyl
8/5/82
60
56
65
47
48
40
40
46
51
60
513

89
94
62
65
CMA-A 2040
Spike
level
Total sample
concentration
CMA-A 2050
Blind quantitat ion
CMA-A 2060Q standard
Spike Total sample Spike Total sample Spike Total samp
level concentration level concentration level eonceiitr_at
8/5/82
20
10
10
15
17
19
25
45
49
42
252





80
58
75
55
58
48
58
82
93
110
717
Recovery
79
93
56
57
49
25
25
36
42
46
62
110
120
100
615
(%) of




8/6/82
92 100
58 51
39 51
43 75
64 86
61 95
87 130
100 230
130 250
140 210
814 1,280
surrogate compounds
76
84
41
48
8/6/82 fi
100 82 140
69 42 53
44 42 87
50 61 110
73 70 140
67 77 160
87 100 340
110 180 560
140 200 530
140 170 430
920 1,020 2,550

93
93
53
64
ile Spike
ion level
i/6/82
184
94
94
137
157
173
234
414
450
369
2,306

88
88
78
79
      a  Concentration in ng/ml  rather than pg/g since this sample was prepared by dilution of stock solutions of standards by QC personnel.

      b  302 Dal.tons used for quantisation.

-------
                  TABLE 32.  CORRECTED PCB CONCENTRATION (pg/g) OF SPIKED CHA-A SAtlPLES DETERMINED BY SURROGATE RECOVERY  CORRECTION
PCB homolog
CGC/EIHS analysis date
Honochlorobiphenyl
Dichlorobiphenyl
Trichl.orobiphenyl
Tetrachlorobiphenyl
Penta ch lorobipheriy 1
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB
CHA-A 2030
Total sample
concentration
8/5/82
67
63
70
50
51
43
64
74
81
91
650

Spike
cone.

20
10
10
15
17
19
25
45
49
42
250
CHA-A 2040
Total sample
concentration
8/5/82
100
74
80
58
63
52
100
150
170
180
1,030

Spike
cone.

49
25
25
36
42
46
62
110
120
100
620
CHA-A 2050
Total sample
concentration
8/6/82
120
76
46
52
77
72
210
250
330
280
1,510

Spike
cone.

100
51
51
75
86
95
130
230
250
210
1,280
CHA-A 2060
Total sample
eoncentration
8/6/82
110
74
47
53
78
72
160
210
270
220
1,290
Blind quantitation
standard 2070Q
Spike
cone.

82
42
42
61
70
77
100
180
200
170
1,020
Total sample
concentration
8/6/82
160
60
99
130
. 160
190
430
720
680
540
3,190
Spike
cone.

184
94
94
137
157
173
234
414
450
369
2,310
Concentration in ng/ml rather than pg/g since this sample was a blind quantitation  sample

-------
  TABLE 33.   PCB CONCENTRATION (pg/g) OF CMA-A SAMPLES TREATED WITH DIFFERENT
               CLEANUP PROCEDURES (INTERNAL STANDARD QUANTITATION)

CMA-A 2090 CMA-A 2110 CMA-A 2130
PCB homolog acid cleanup Florisil cleanup alcoholic KOH cleanup
CGC/EIMS analysis date
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorpbiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB

13C6-monochlorobiphenyl
13Ci2-tetrachlorobiphenyl
13Ci2~octachlorobiphenyl
13C12~decachlorobiphenyl
8/9/82
NDa
4.4
0.4
25
19
7.9
5.9
2.4
38
16
119
Recovery
74
0
115
173
8/9/82
4.4
14
31
18
17
5.6
2.2
6.0
2.4
9.5
110
(%) of surrogate compounds
8
0
97
129
8/9/82
31
44
44
25
20
6.3
3.8
2.6
2.6
6.4
186

145
367
110
64

a  ND = not detected.

b  302 Daltons used for quantitation.
                                   54

-------
           TABLE 34.  PCB CONCENTRATION (M8/g) OF CMA-A SAMPLES TREATED WITH
	VARIOUS CLEANUP PROCEDURES (SURROGATE COMPOUND CORRECTED)	

                            CMA-A 2090         CMA-A 2110             CMA-A 2130
PCB homolog                acid cleanup     Florisil cleanup     alcoholic KOH cleanup
CGC/EIMS analysis date
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB
8/9/82
NDa
30
0.3 (0.2)b
17 (11)
13 (8.3)
5.3 (3.5)
2.6
1.1
17
3.9
90 (78)
8/9/82
28
86
200 (16)
110 (9.3)
110 (8.9)
3.5 (2.2)
1.2
3.1
1.2
3.1
546 (159)
8/9/82
11
15
15 (20)
8.4 (11)
6.8 (9.0)
2.9 (2.9)
1.8
1.2
1.2
4.2
68 (77)

a  ND = not detected.

b  13C!2-tetrachlorobiphenyl was not quantifiable due to interferences.  The values
     reported were calculated using ^Cg-monochlorobiphenyl.  Values in parenthesis
     were calculated using 13C12-octachlorobiphenyl.
                                       55

-------
DCMA Pigment Samples--
     Eight DCMA pigment samples were analyzed following the preparation de-
scribed in the experimental section (Table 15).  The results are presented in
Table 35.  The diarylide yellow pigment (DCMA-1) was analyzed in duplicate
and as a blind spike supplied by the MRI quality control department.  This
sample is reported to contain 3,3'-dichlorobiphenyl at levels of approximately
70 |Jg/g (Dry Colors Manufacturing Association, 1981).  No analyte or surrogate
PCBs were detected in the duplicate 1-g samples of the pigment and a known
spike of the sample.  The lack of detected PCBs indicates a loss of analytes
in the sample preparation.  The CGC/EIMS analysis of a sample of the yellow
pigment spiked by MRI quality control personnel yielded an uncorrected con-
centration of 76 M8/8 °f 3,3'-dichlorobiphenyl based on the internal standard
quantitation and a corrected concentration of 63 M8/8> based on 12070 recovery
of the 13C6~4-monochlorobiphenyl surrogate.  The level of the 3,3'-dichloro-
biphenyl added by the QC personnel was reported to be 60 P8/8-  Hence, the
total dichlorobiphenyl concentration should have been approximatey 130 |Jg/g
(70 ng/g endogenous plus 60 (Jg/g added).

     The phthalocyanine green pigment (DCMA-4) was also analyzed in duplicate
following dissolution and fractionation with a Florisil column.  This pigment
reportedly contains only decachlorobiphenyl at approximately 40 (Jg/g based on
the results of the DCMA round robin study (Dry Color Manufacturing Associa-
tion, 1981).  Our analysis of duplicate samples yielded uncorrected concen-
tration levels of 24 and 27 (Jg/g of decachlorobiphenyl by the internal quanti-
tation method.  The corrected concentration for both samples was 13 (Jg/g with
recovery of the 13Ce-decachlorobiphenyl surrogate at 190 and 210%.

     Phthalocyanine blue (DCMA-8) was also analyzed as a single sample.
Pentachloro- and hexachlorobiphenyls were detected but the concentrations
were below the quantitation limits for that particular day.  The total PCB
concentration of this pigment, as discussed in the results of the DCMA round
robin (1981), is reported to be 90 |Jg/g.

     The DCMA pigment sample analyses did not produce valid results.  These
data suggest that further development or validation of extraction/cleanup pro-
cedure would be necessary to provide acceptable PCB analyses of these samples.
All of the blank samples (2001, 2080, and 2100) analyzed along with the DCMA
samples were found to contain no detectable PCBs.

DISCUSSION

     The determination of PCBs is a complex problem.  The inaccessability of
standards for all 209 congeners has traditionally been circumvented by the
use of commercial mixtures (e.g., Aroclors) as standards.  Quantitation has
often been addressed in terms of relating the analyte to an Aroclor standard
to give a "total PCB" concentration.  Determination of PCBs synthesized as
by-products in commercial products or product waste presents three special
problems:  (a) the analyte does not generally resemble a commercial PCB mix-
ture, so quantitation against Aroclor standards would be incorrect; (b) the
matrix often contains high concentrations of other chlorinated organics which
are not easily separated during a cleanup procedure and which interfere with
the qualitative and quantitative analysis; and (c) the matrix is undefined
and can include gases, liquids, or solids of any purity and complexity.

                                    56

-------
             TABLE 35.   RECOVERY (%)  OF CARBON-13 LABELED SURROGATE COMPOUNDS  FROM DIARYLIDE  YELLOW
                                    AND PHTHALOCYANINE BLUE AND GREEN PIGMENTS

PCB
surrogate
13C6-Monochlorobiphenyl
13C12-Tetrachlorobiphenyl
13C12-Octachlorobiphenyl
13C12-Decachlorobiphenyl
DCMA-1
21403
ND8
ND
ND
ND
DCMA-1
21503
ND
ND
ND
ND
DCMArl
216?
ND
ND
ND
ND
DCMA-1
2170QC
120
ND
200
250
DCMA^A
2175d
ND
ND
120
190
DCMA-:4
2180
ND
ND
107
210
DCMA-8
21906
ND
94
92
150
DCMA-8
2200Q
12
52
71
77

a  Samples 2140 and 2150 are duplicates prepared by the DCMA-B method.

b  Sample 2160 was spiked with 50 pg/g of 3,3'-dichlorobiphenyl and prepared by the DCMA-B method.

c  Sample 2170Q was spiked by MRI quality control personnel with 3,3'-dichlorobiphenyl and was prepared
     by the DCMA-B method.

d  Samples 2175 and 2180 are duplicates prepared by the DCMA-B method.

e  Sample 2180 was prepared by the DCMA-A method.

f  Sample 2200Q was weighed by MRI quality control personnel.   An unknown mass of sample was supplied for
     preparation by the DCMA-A method.

g  The four surrogate compounds were added but not detected.

-------
     In this situation, analytical methods require a different philosophy
than the classic approach for a single analyte in a defined matrix where all
steps, reagents, and apparatus are specified.  The method proposed here leaves
many of the analytical steps to the discretion of the analyst while ensuring
the reliability of the results with a strong quality control program.  Thus,
an analyst familiar with general analytical techniques for a product, may read-
ily adapt in-house extraction/cleanup procedures to incidental PCB analysis.
Even when the recoveries are not optimized, the l3C-labeled surrogate recov-
eries will mimic those of the analyte PCBs.  As long as the 13C recovery sur-
rogates are thoroughly incorporated, their recoveries can be used to derive
corrected analyte PCB concentrations.

     Several of the method validation analyses presented above, especially
Tables 33 and 35, illustrate the importance of the recovery surrogates in QC.
The techniques employed are common methods validated for PCB analysis by other
laboratories without the 13C-surrogate data.  Analyses of this type may have
been used by a testing laboratory and erroneous results reported.

     The complexity of the matrix and the high probability of chlorinated or-
ganic interferents precluded the use of GC/ECD.  The best available technique
is GC/EIMS.  During the validation work presented above, the anticipated dif-
ficulty of qualitatitve and quantitative data interpretation was confirmed.
In addition to the inherent problems resulting from extrapolation from a stan-
dard to several analytes, interpretation of the complex peak clusters is a
tedious, subjective, and error-prone process.  The volume of data for one
sample is staggering; for sample 2110, 286 peaks were identified and inte-
grated in the PCB mass chromatograms as shown in Figures 6 through 16.  Of
these, 58 peaks met the qualitative criteria and were identified as PCBs.
Clearly different analysts will obtain different results for those peaks
which marginally fit the qualitative criteria.  This very high data density
relative to other common GC/MS analyses has a much higher potential for error
and mistakes.  In addition it should be noted that, for many of the samples
analyzed in this study, the data interpretation is more time-consuming than
the rest of the analytical process.

     The integration methods are also prone to error.  Integration is always
conducted interactively with the mass spectrometric data system, either man-
ually or automatically.  The selection of baseline criteria, background sensi-
tivity, integration method (valley-to-valley, baseline-to-baseline, etc.),
and retention window all affect automatic quantitation.  The position of the
cursor and integration method affect the manual quantitation results.

     The day-to-day instrumental variability with quadrupole systems also ap-
pears to adversely affect data quality, despite tight calibration specifica-
tions.  The magnitude of this error soruce should be further documented.

     The above discussion presents some of our understanding of some of the
major problems with analysis for by-product PCBs.  Further work will be de-
voted to characterizing and reducing these problem areas.  Even with forsee-
able improvements in the method, the data for by-product PCBs in many com-
mercial product and product waste samples will exhibit low precision and
accuracy.


                                    58

-------
VO
BIG                                              DATA: 4901IIO$J5 01
08/09/82 16:20:00                                  CALI: MIDCAIJWOVI fl4
SAIIPLE: SAMPLE 82110 QIA-A FLORISIL  1/IOOIL  IUIIIU
RANGE: G   1.1750  LABEL: II  0.  4.0  GUAM: A  9. 1.0 RASE: U 2«.  3

                                        993
                                                                                                    SCAIIS    1 in 1750
                          200
                          3:19
           6:20
 800
12:40
1880
I5:'JO
1200
19:08.
1409
22:10
1660
25:2«
SCA1I
TIHF.
            Figure 6.   Reconstructed  ion  chroraatogram  for  SIM  analysis  of  the  CFA-A sample  No.  2110.

-------
37.9-1
188
29. H
              MASS CIIKOIIATOGRAIIS                               DATA; 190III09V5 Bl
              08/09/82 16:20:00                                CAM: IIIDTALIMWI fl1
              SAMPLE:  SAMPLE S2M0 CIIA-A FIOR1S1L  I/IOIUL  IULHU
              RANGE: G  1.1750  LADEI.: H 9, 4.0  I3UAH: A  9.  1.0  BASE: U  29.  3
                                                                 832
                                   SCAHS  709 TO  900
                                                  916170.
   im
   11:05
12:40
 850
13:27
 900  sati
14:15 THIJ-:
 Figure  7.   SIM ion  plots  for  monochlorobiphenyls  (188 and  190 Daltons) and  the    C^-

    monochlorobiphenyl surrogate  (194 Daltons) in CMA-A sample No.  2110.

-------
100.8-!
 222.
 45.3-1
 224 .
                MASS CIIBOIIATOGRA11S                                  DATA: 1901II99V5 Bl
                88/09/82 16:20:00                                   CALI: IIIDCAUHW'Jl Bl
                SAMPLE: SAIfPLE 82110 dlA-A FLORISIL  t/IODfL  1UI.IHJ
                RAIIGE: G   1.1759  LABEL:  II 8. 4.0  QtJAII: A  0.  1.9  BASE: U 20.  3
SCAMS 700 TO 1100
                                                                                1080
                                                                                15:59
                                                                                                       1099
                154001^1.
                 222.RP6
                *  0.5C3
                                                                                                             6983P8.
                 224.607
                *  9.SC9
 1059
 16:37
1109  SCAN
17:25 Tllll-
    Figure 8.   SIM  ion  plots for dichlorobiphenyls  (222  and  224 Daltons)  in CMA-A sample

       No.  2110.

-------
NJ
                             MASS CHBOHATOGRAHS                                  DATA:  490IH09V5 til
                             08/89/82 16:20:80                                   CAL1:  tilDTALII09VM M
                             SAIIFLE: SAIIPI.E A2II0 QIA-A FLORISIL  1/I0DIL  IULIIU
                             RAIIGE: G   I.I750  LABEL:  \\ 9. 4.9  01)All: A  0. 1.0  BASE: U 29.  3
                                                       SCAIIS 850 TO
             180.0
              256 .
              89.6-
              258
                                                                                                   1093
                                                          972.	99?	
                                                                         1017
                                                                                                   1093
                                                                         1PI6
                 850
                13:27
 900
14:15
 950
15:92
=»f—f	r—-r-	i	1

 1000              1050
 15:50             16:37
1100
17:25
                                                                       1156120.
                                                                        256.077
                                                                       *  0.500
                                                                       I304570.
                                                                        258.077
                                                                       *  0.599
1150 SCAH
18:12
              Figure 9.   SIM ion plots for trichlorobiphenyls  (256  and 258 Daltons)  in  CMA-A  sample  No.  2110.

-------
 71.3n
  299 _
               MASS QIBOIIMOCRMIS                               DATA: 499III99V5 81
               98/99/82 16:20:90                                CA1.|. HIBCALIW9tfl 01
               SAMPLE: SAMPLE 82119 CIIA-A FLORISIL  1/I0DIL  1ULUU
               BAHGE: G   1.1759  LABEL:  II 9. 4.9  GUAII: A  9.  1.9  BASE: U 29.  3
                                                          1222
                                                  SCAHS 1959 TO 1350
 109.9-1
  292 .
 39.7-1
 298 _
 62.7-1
 394 .
    1659
    16:37
1158
18:12
1269
19:99
1259
19:47
1399
29:35
                                                                  217296.
                                                                  :>39.W»7
                                                                  .'-  0.509
                                                                                                      316621.
                                                                  292.987
                                                                    0.590
                                                                   137728.
                                                                  298.989
                                                                    9.599
                                                                   217344.
                                                                   304.991
                                                                    9.599
1350  SCAM
21:22 TIME
Figure 10.   SIM ion plots for  tetrachlorobiphenyls (290 and  292  Daltons),  3,3',A,A'-tetrachloro-

  biphenyl-dg  (298  Daltons), and  the 13C;L2-tetrachlorobiphenyl surrogate (304 Daltons)  in  CMA-A

  sample  No. 2110.

-------
109.0-1
326 .
81.3-1
328
MASS CIIIiOIIATOGRAIIS                                  DATA:  190III09U5 HI669
08/89/82  16:28:80                                  CALI:  IIIDCALII99VI 11
SAMPLE: SAMPLE 82II8 CIIA A FLORISIL  l/IODIL  1UIIIU
RAHGE: C   1,1758 LABEL: II  8. 4.8  OIJAII: A  8. 1.8  BASE: U 20,   3

       12 7
                                                                                           SCAIIS !208 TO 15QO
   1200
   19:89
                                                                        15897R.
                                                                                             326.897
                                                                                            *  8.588
                                                                                             129288.
                                                                                             328.898
                                                                                            *  8.588
     1258
     19:17
1300
20:35
14Q8
22:18
1158
22:57
1563 SCAH
23:15 TIIIE
    Figure 11.   SIM  ion  plots for pentachlorobiphenyls  (326  and 328 Daltons)  in  CMA-A  sample

       No.  2110.

-------
 97.4-1
               »ASS aiROIIATOCRAm                                 DATA: 49eiII0flV5 81
               68/09/82 16:29:89                                  CALI: IIIDCALII03!M B1
               SAIIFLE:  SAIIPLE H2II9 CIIA-A FLOItlSIL  l/IODIL  IULUU
               RANGE: 6   1.I759 LADEL: II  9. 4.9  OUAII: A  9.  1.9  CASE:  U 29.  3
                                               SCANS 1259 TO 1509
 369 .
180.0-1
 362 .
    1259
    19:47
1359
21:22
1499
22:19
1459
22:57
                                                                150784.
                                                               309.188
                                                               *  9.599
                                                                154889.
                                                               302.198
                                                                 9.599
1593 SCAM
23:15 TIME
  Figure  12.   SIM ion plots  of hexachlorobiphenyls  (360 and 362 Daltons)  in CFA-A sample No.  2110,

-------
OX
                80.8-1
                394
                 J.O-,
                              IttSS OIR011ATOGRAIIS                                 DMA: 4901H00V5 81         S€MIS 1358 TO I5TO
                              08/09/82 16:20=00                                  CAll: IIIDCALIWWI D1
                              SAIIFLE: S.MIPLE I2H0 CHA-A FLORISIL  1/I0DIL  lULIIU
                              RAHGE: G   1,1750  LABEL: II  0.  4.0  (IIJ.AII: A  0. I.0 ItASE: U 20.  3
                                 1379
                                                                                                    1516
     45588.
                                                                                                                         391.118
                                                                                                                           0.508
     56384.
                                                                                                                         396.118
                                                                                                                          0.590
                  1359
                  21:22
1559 SCAH
2-1-.32 THIE
                 Figure  13.   SIM ion plots  of heptachlorobiphenyls  (394  and  396  Daltons)  in CMA-A  sample No.  2110.

-------
 70.3-1
 428 _
               IIASS aiBOIIATOGRAIIS                                DATAt 490IHB9V5 Of
               08/B3/82 16:20:09                                 CALIs HIDCAUieOVI HI
               SAJIPLEt SA11FLE 82110 CIIA-A FLOP.ISIL  I/1601L  1ULUU
               RAIIGE: G   1.1750  LABEL: II 0. 4.0  CUAII: A  0. 1.0  BASE: 0 20.  3
                                              1532
                                                         SCAHS 1440 TO 1650
 98.9-1
 430 _
1W.0-
 442

                                                                          29728.
                                                                         4:>8. 128
                                                                          11856.
                                                                         438.129
                                                                        >.  0.500
                                                                          42304.
                                                                 163<,     442.132
                                                                 "*•-     *  0.509
        1450
        22:57
1509
23:45
1550
24:32
1609
25:20
1C50 SCAN
20:07 TIME
  Figure  14.   SIN  ion  plots of  octachlorobiphenyls  (428  and  430  Daltons) and the    C^2~octa"

     .chlorobiphenyl surrogate  (442 Daltons)  in CMA-A  sample No.  2llO.

-------
OS
00
                          MASS dlKOIIATOCRAIIS                                 DATA: 4901II09V5 01
                          08/09/82 16:20:60                                  CAI.I: MIDCALIIOWI HI
                          SAMPLE: SAMPLE 82110 CMA-A FLORISJL  1/UMHL  1ULIIU
                          RAIIGE: C   1,1759  LABEL: II  0. 4.0  QUAD: A  0.  1.0  HASH: U 20.  3
                                                         SCAII3 1559 TO 1650
           89.5-,
           461
                                                                                                  1631
                                                                           24256.
                                                                          164.139
                                                                            0.500
           100.0-,
                                        1576
           466
                                                                           27104.
                                                                          466.139
                                                                         *  0.500
                        1560
                        24t42
1580
25:01
1680
25:20
1620
25:39
1610
25:58
SCAN
TIME
           Figure 15.   SIM  ion  plots of nonachlorobiphenyl (464  and  466  Daltons) in CMA-A  sample  No.  2110.

-------
 89.8-1
 498
189.8-1
 509 .
 49.7-,
 518.
               I1ASS C1IB01UTOGRAIIS                                DATA:  490III09V5 BI669
               88/09/82 16:26:09                                 CALI:  IIIDCALIIOTVI til
               SAIIFLEt SAMPLE 121 IB OIA-A FLOBISIL  I/IODIL  IULIHJ
               BAHGE: G   1.1758  LABEL: H 8. 4.0  QUAth A  8. 1.8  BASE: U 29.  3
                                        1669
                                                                 SCAIIS 1659 TO I7G3
   1650
   26:87
1668
26:17
                                                                                  79232.
                                                                                   08.119
                                                                                   0.509
                                                                                             1696
                                        1669
                                                                                        tgg^^L.
                                                                                   88I92.
                                                                                  499.589
                                                                                 * 0.589
                                                                      1684
                                         1669
                                                                                             1696

                                                                                            -/^
                                                                                   43840.
                                                                                  5^9.508
                                                                                 *  0.588
                                                                                                 IC99
1678
26:26
16C0
26:36
1698
26:45
1700 SCAH
2P:K> TIME
                                                                                                 13,
Figure 16.   SIM ion plots  of decachlorobiphenyl (498 and 500 Daltons) and the    C12~

    decachlorobiphenyl  (510 Daltons) in CltA-A sample No.  2110.'

-------
                                  SECTION 5

                                 REFERENCES

Dry Color Manufacturers Association.  1981.  An analytical procedure for the
determination of polychlorinated biphenyls in dry phthalocyanine blue,
phthalocyanine green, and diarylide yellow pigments.  1117 North 19th Street,
Arlington, VA  22209.

Erickson MD, Stanley JS.  1982.  Midwest Research Institute.  Methods of
analysis for incidentally generated PCBs literature review and preliminary
recommendations.  Draft interim report no. 1.  Washington, DC:  Office of
Toxic Substances, U.S. Environmental Protection Agency.   Contract 68-01-5915.

Haile CL, Baladi E.  1977.  Midwest Research Institute.   Methods for deter-
mining the total polychlorinated biphenyl emissions from incineration and
capacitor and transformer filling plants.  Washington, DC:  U.S. Environ-
mental Protection Agency.  Contract 68-02-1780.  EPA 600/4-77-048.

Pittaway AR, Horner TW.  1982.  Heiden, Pittaway Associates.  Statistical
analysis of data from a round robin experiment on PCB samples.  Washington,
DC:  Chemical Manufacturers Association report.

Roth RW, Keys JR, Chien DHT, et al.  1982.  Midwest Research Institute.
Methods of analysis for incidentally generated PCBs—synthesis of 13C-PCB
surrogates.  Draft interim report no. 3.  Office of Toxic Substances, U.S.
Environmental Protection Agency.  Contract 68-01-5915.

Stanley JS, Erickson MD.  1982.  Midwest Research Institute.  Peer review and
authors' replies to 'methods of analysis for incidentally generated PCBs--
literature review and preliminary recommendations.'  Draft interim report no.
2.  Washington, DC:  Office of Toxic Substances, U.S. Environmental Protection
Agency.  Contract 68-01-5915.

USEPA.  1979a (December 3).  U.S. Environmental Protection Agency.   Base/
neutrals, acids, and pesticides—method 605.  44 FR 69540.

USEPA.  1979b (December 3).  U.S. Environmental Protection Agency.   Organo-
chlorine pesticides and PCBs--method 608.  44 FR 69501.

USEPA.  1982 (June 8).  U.S. Environmental Protection Agency.  Polychlorinated
biphenyls (PCBs); manufacture, processing, distribution, and use in closed
and controlled waste manufacturing processes.  FR 74 24976.
                                    70

-------
                APPENDIX A






SUPPLEMENTARY GC/EIMS DATA ON PCS CONGENERS
                  A-l

-------
     The following data support the method validation section for gas
chromatography/electron impact mass spectrometry (GC/EIMS) of polychlorinated
biphenyls (PCB).   Table A-l lists the average relative response factors (RRF)
for the 77 commercially available PCB congeners determined as four replicates.
Table A-2 presents results of the Student's t-test used to determine the sig-
nificance of differences for average RRFs for PCB homologs measured on a
single day versus multiple days.  The data in Table A-2 indicate that only
the average RRFs for the heptachlorobiphenyl homolog are significantly dif-
ferent.

     Table A-3 presents the results of the Student's t-test used to determine
the significance of differences for the average RRFs for the PCB homologs de-
termined with the quadrupole and magnetic sector mass spectrometers.  All 77
PCB congeners were determined in a single day for each of the instrument stud-
ies.  This comparison indicates that the average RRF values are significantly
different, which was expected.  However, the relative standard deviations are
not significantly different, indicating that the selection of the calibration
standards is appropriate.  These conclusions are discussed more fully in the
text.

     Table A-4 presents results of the Student's t-test used to determine
significance of differences for the RRFs for the 11 congeners in Solution
No. 1, which was analyzed daily.  An example of the data generated for multi-
ple analysis of Solution No. 1 is presented in Figures 1 to 23.  This infor-
mation includes a capillary GC/EIMS chromatogram of Solution No. 1, the mass
spectra of each component in this solution, and a graphic illustration of the
distribution of several measurements of each congener about the average re-
sponse factor.  It should be noted that the standard deviation and relative
standard deviation presented in these plots are different from that reported
in the text due to calculation of the standard deviation using N weighting
rather than the correct N-l weighting.  All other standard deviations reported
in this document are based on the N-l weighting.

     The relative retention times of the 77 PCB congeners with respect to
3,3',4,4'-tetrachlorobiphenyl-d6 determined with the Finnigan 4023 quadrupole
and the Varian MAT 311A mass spectrometers are presented in Table A-5.  A
relative retention time unit of 0.01 (10 sec) is required for resolution of
two specific congeners based on the gas chromatography parameters used to gen-
erate these numbers.
                                   A-2

-------
TABLE A-l.  RELATIVE RESPONSE FACTORS FOR COMMERCIALLY
        AVAILABLE PCS CONGENERS (QUADRUPOLE)

Congener
no.
1
2
3
4
5
7
8
9
10
11
12
14
15
18
21
24
26
28
29
30
31
33
40
44
47
49
50
52
53
54
61
65
66
69
70
72
75
77
Degree of
chlorination
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
Average relative
response factor
, 4.073
2.951
2.969
1.232
1.959
2.008
2.049
2.148
1.880
3.073
1.929
2.083
1.909
1.104
1.586
1.051
1.714
. 1.587
2.195
1.526
1.706
1.688
0.597
0.712
1.062
0.831
0.957
0.732
0.750
0.958
0.975
1.086
1.139
1.058
1.091
1 0.980
1.185
1.095
Standard
deviation
0.118
0.056
0.028
0.008
0.035
0.027
0.023
0.061
0.031
0.073
0.036
0.098
0.089
0.012
0.018
0.033
0.013
0.028
0.048
0.067
0.024
0.031
0.013
0.007
0.059
0.019
0.025
0.011
0.008
0.013
0.069
0.022
0.068
0.012
0.050
* 0.048
0.061
0.050
Coefficient of
variation (%)
2.905
1.894
0.956
0.646
1.803
1.366
1.134
2.846
1.658
2.363
1.877
4.702
4.686
1.089
1.110
3.105
0.731
1.733
2.188
4.418
1.409
1.863
2.152
0.946
5.591
2.245
2.574
1.504
1.006
1.344
7.094
1.994
5.966
1.110
4.548
4.870
5.113
4.595
                             (continued)
                       A-3

-------
                         TABLE A-l (continued)

Congener
no.
87
88
93
97
100
101
103
104
115
116
119
121
128
129
136
137
138
139
141
143
151
153
154
154
155
156
171
181
183
185
195
198
200
202
204
206
207
208
209
Degree of
chlorination
5
5
5
5
5
5
5
5
5
5
5
5
6
6
6
6
6
6
6
6
6
6
6
6
6
6
7
7
7
7
8
8
8
8
8
9
9
9
10
Average relative
response factor
0.617
0.611
0.574
0.719
0.727
0.653
0.566
0.824
0.853
0.785
0.762
0.948
0.499
0.431
0.689
0.533
0.433
0.462
0.419
0.490
0.473
0.549
0.221
0.511
0.587
0.599
0.346
0.383
0.380
0.336
0.263
0.262
0.301
0.250
0.221
0.193
0.237
0.259
0.213
Standard
deviation
0.011
0.005
0.010
0.008
0.003
0.004
0.009
0.025
0.061
0.013
0.022
0.020
0.005
0.004
0.016
0.008
0.008
0.026
0.010
0.005
0.013
0.050
0.001
0.010
0.011
0.044
0.002
0.009
0.010
0.006
0.003
0.008
0.007
0.007
0.007
0.003
0.008
0.003
0.006
Coefficient of
variation (%)
1.710
0.744
1.677
1.139
0.428
0.538
1.627
3.048
7.146
1.654
2.911
2.127
1.093
0.813
2.336
1.582
1.946
5.686
2.353
0.986
2.826
9.101
0.570
2.039
1.828
7.431
0.640
2.379
2.501
1.729
1.184
2.887
2.392
2.663
3.200
1.723
3.547
1.315
2.837

Relative to 3,3*,4,4'-tetrachlorobiphenyl-de-  All relative response
  factors were calculated as the average of four replicate measurements
  made on the same day.
                                A-4

-------
                   TABLE A-2.  STUDENT'S TWO-SIDED t-TEST TO DETERMINE SIGNIFICANT DIFFERENCES BETWEEN
                       QUADRUPOLE RESPONSE FACTORS CALCULATED ON THE SAME DAY VERSUS MULTIPLE DAYS
Ul

Number of
PCB homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
isomers
3
10
9
16
12
13
4
6
3
1
Average RRF
from
replicate
measurements
3.331
2.027
1.573
0.950
0.720
0.513
0.361
0.253
0.229
0.213
Standard
deviation
0.643
0.447
0.341
0.175
0.120
0.078
0.024
0.030
0.034
0.006
Average RRF
from
single ,
measurement
2.739
2.048
1.592
0.946
0.725
0.500
0.308
- 0.224
0.188
0.179
Standard
deviation
0.254
0.322
0.289
0.189
0.127
0.096
0.025
0.039
0.030

Significant
t-Statistic at
1.478
-0.119
-0.131
0.0618
-0.1085
0.377
3.119
1.398
1.|J91

95% level?
No
No
No,
No
No
No
Yes
No
Na


       a  Four replicate measurements of the RRF were made for each isomer.   For example,  the three monochloro-
            biphenyl isomers were measured four times each.  Hence, the average RRF and standard deviation were
            calculated from 12 distinct values.

       b  A single measurement for each of the 77 PCB congeners was completed in a single  day.  Hence, the
            average RRF reported is the average of one measured RRF for each isomer within a homolog.  For
            example, the average RRF and standard deviation reported for the monochlorobiphenyl was calculated
            from three distinct values.

       c  Single measurement.

       d  Cannot test significance of difference between single measurements.

-------
        TABLE A-3.   COMPARISON OF THE AVERAGE RELATIVE RESPONSE FACTORS (RRF) DETERMINED WITH QUADRUPOLE
                    (FINNIGAN 4023)  AND MAGNETIC SECTOR (VARIAN MAT 311A)  MASS SPECTROMETERS'"

Finnigan 4023
quadrupole MS
PCB homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
Number of
isomers
3
10
9
16
12
13
4
6
3
1

RRF
2.739
2.038
1.592
0.946
0.725
0.500
0.308
0.224
0.188
0.179
Standard
deviation
0.250
0.32
0.29
0.19
0.13
0.10
0.025
0.04
O.g3
Varian MAT 311A
magnetic
sector MS

RRF
2.329
1.663
1.167
0.902
0.780
0.640
0.497
0.463
0.467
0.586
Standard
deviation
0.199
0.229
0.248
0.130
0.136
0.124
0.060
0.071
O.J05

RRFs significantly
different at the ,
95% confidence level
No
Yes
Yes
No
No
Yes
Yes
Yes
Yes
e
Variances significantly
different at the
95% confidence level
No
No
No
No
No
No
No
No
No
_e

a  The RRF and standard deviation reported in this table for the quadrupole and magnetic sector mass spectrometers
     were determined as single measurements of all congeners in a single day with each instrument.
b  Student's two-sided t-test was used to determine significant differences of the RRFs.

c  An F-test was used to determine significant differences of the standard deviations, where
     F = (std dev!)2/(std dev2)2 with (n-1, n-1) degrees of freedom.

d  Single measurement.

e  Cannot test significance of difference between single measurements.

-------
    TABLE A-4.  STUDENT'S TWO-SIDED t-TEST TO DETERMINE SIGNIFICANT DIFFERENCES OF THE AVERAGE RELATIVE
                      RESPONSE FACTOR (RRF) FOR SOLUTION NO. 1 FOR REPLICATE ANALYSIS
                          ON A SINGLE DAY VERSUS SINGLE ANALYSES ON MULTIPLE DAYS

PCB
congener
no.
1
11
29
47
121
136 -
181
195
207
209
Replicate analyses
on single day

RRF
4.073
3.073
2.195
1.062
0.948
0.689
0.383
0.263
0.237
0.213
Standard
deviation
0.118
0.073
0.048
0.059
0.020
0.016
0.009
0.003
0.008
0.006
Single analyses,
on multiple days

RRF
3.241
2.538
1.899
1.015
0.959
0.683 -
0.374
0.275
0.269
0.230
Standard
deviation
0.201
0.161
0.100
0.059
0.043
0.058
0.035
0.028
0.032
0.027

t-Statistic
7.468
6.204
5.483
1.268
-0.479
0.186
0.662
-1.137
-2.479
-1.599
Significant differences
of RRF at
confidence
Yes
Yes
Yes
No
No
No,
No
No
Yes
No
95%
limit?











a  The RRF and standard deviations were calculated from four replicate measurements completed in the same
     day.
b  The RRF and standard deviatons were calculated from seven single measurements from seven different days.

-------
TABLE A-5.  RELATIVE RETENTION TIMES (RRT) OF 77 COMMERCIALLY AVAILABLE
     PCB CONGENERS MEASURED VERSUS 3,3*4,4'-TETRACHLOROBIPHENYL-d6
         DETERMINED WITH MAGNETIC SECTOR (VARIAN MAT 311A) AND
             QUADRUPOLE (FINNIGAN 4023) MASS SPECTROMETERS

RRT
PCB congener no.
Monochloro-
1
2
3

Dichloro-
4
5
7
8
9
10
11
12
14
15

Trichloro-
18
21
24
26
28
29
30
31
33

Tetrachloro-
40
44
47
49
50
52
53
54
61
65
66
69
70
72
75
77
311A

0.403
0.481
0.474


0.518
0.598
0.559
0.590
0.563
0.521
0.649
0.660
0.616
0.677


0.665
0.762
0.685
0.729
0.745
0.719
0.641
0.741
0.760


0.870
0.838
0.814
0.811
0.746
0.804
0.763
0.720
0.898
0.822
0.905
0.800
0.880
0.853
0.816
1.002
4023

0.425
0.490
0.499


0.536
0.606
0.579
0.606
0.577
0.534
0.660
0.671
0.628
0.681


0.678
0.767
0.694
0.738
0.753
0.728
0.653
0.752
0.769


0.875
0.843
0.819
0.817
0.751
0.810
0.773
0.731
0.898
0.826
0.908
0.807
0.904
0.856
0.821
1.003
PCB congener no.
Pentachloro-
87
88
93
97
100
101
103
104
105
116
119
121

Hexachloro-
128
129
136
137
138
139
141
143
151
153
154
155
156

Heptachloro-
171
181
183
185

Octachloro-
194
195
198
200
202
204



RRT
311A

0.979
0.913
0.907
0.976
0.878
0.945
0.870
0.829
0.988
0.985
0.964
0.911


1.163
1.128
0.994
1.118
1.108
1.037
1.096
1.050
1.020
1.074
1.002
0.929
1.194


1.189
1.178
1.154
1.166


1.355
1.326
1.275
1.203
1.194
1.209



4023

0.978
0.915
0.908
0.979
0.884
0.945
0.874
0.836
0.987
0.986
0.965
0.914


1.156
1.127
0.996
1.115
1.103
1.038
1.093
1.051
1.021
1.073
1.004
0.931
1.188


1.187
1.174
1.148
1.161


1.351
1.317
1.265
1.199
1.188
1.203



                                             (continued)
                                   A-8

-------
                           TABLE A-5 (continued)
                         RRT
                                           RRT
PCB congener no.    311A     4023     PCB congener no.    311A     4023
Nonachloro-
    206
    207
    208
1.414
1.336
1.319
1.399
1.330
1.318
Decachloro-
    209
1.453    1.440
                                     A-9

-------
        0:33:00
SAMPLE:  PCB IHXTUKE—SOLUTIOII III
CO!!»S.:  -1600 HIV
RAIitiE: C  1.1?CO
                                      g»: CSS??!1765
                                     IIJL [||J
TO
TO
I2fl0
I2CO
                        10-6 70EV .2IIA nD!i-HI( 10-2II-325-C/ "OII-COI"
                         LADEL: I!  0. 1.0 U.UAMs A 0. I.O J  0  BASE: U 20.  3











>
1
r-1
O












-HUM! 1.

^
c1 *-V96
(U
XI
p.
XI >, i-H
o a >>
H 
  • H ~* O 0. i • i 59 c M XI 0 CO 0) 1 ^j « ^ CM * CM .-*. 0. rH C (U XI p. •H ("> O 0 rH XI o CO 4J 0) PL. ?I2 rH C (LI P. •iH O H O rH U Cfl s 1 1 I ^"I ^° In vo" •u -d~ H <^ CO -* " U O 1 in" ** _ CM CM . 1.251 \- —h ^ 599910 C (ll "p. 1.-1M •H Q O 6, 6 '-Nonach] ""1 Zr
    -------
    59.6-
    ii/n
                   G->/jy/:}2 IC: 12:Cfl *  1:5?
                   SAIirLE: PCB IHXIllllE—SOLUTION 111 lUL  lilJ
                   OJiiDS.: -ITOO BIV IO-6 70EV .2IIA Ili'o-ljll IIO--2II-325-IO/ "Kil-CnL
                   I3BIAKCED (S ISO 2H Of)
    DATA: 190in20yi>7 ()29«
    I'M.I: CALF.20UI  m
                                             BA'JK H/K:  I"')
                                             lilC:   295 1.24.
                                                                  •52
                                          12G
                 99
                                                                                      173
                                                                                                  IS8
                                                                                                     189
                                                                                                               9523*'.
                                    129
    .   110
         ICO
                                                                                            100
            Figure  A-2.  Electron impact  ionization mass  spectrum  of  2-monochlorobiphenyl,
    

    -------
    100.0-1
     59.0-
     It'll
        I;A-:S tt'KCTWjii
        WJ/W/Kl Ui:12:00 *  /:'jt
        SAliri.!:: FOB IllXtlll!!:—-SDUMimi III HH. Itll
        owns.: -imo  tnv lo-c ;	
        ElillAliCHD IS 1511 2(1 01)
                DATA: iyi)ll:2«Ui>7  B173
                CAM-. CA1.K20V1  H3
               99
    100
                         ll        l
                         Jlu-iiilu
                              120
                                                         110-211 325- It)/
                                                     C2
    Jt3«
                                                            1DO
    2«»
                                                                                              DAS!: II/'E:  222
                                                                                              UtC:   2GJG!}!).
                                                                                                     ?22
    220
                                                                                                                 C329G.
          Figure A-3.   Electron  impact  ionization  mass spectrum of  3,5-dichlorobiphenyl,
    

    -------
    50.0-
             83
                   II.CS SPECTRI!il
                          KATA:
                          i:AM:
                   05/20/«2  10 :
    -------
    J00.0-1
    50.0-
            83
                   IIASS sr
                   05/20/82  10:12:00 +  f):M
                   SAMPLE: PCD IIIXTIIISE—-5IJJ.IJTIOH Bl IUL II!I
                   CUiJDS.: -l'j'31) B1V 10-0 70liV ,21IA 1)115-IfM ltO-211-325 UV
                   B3WtlCn»  (S I5B 211 OT»
                                                                      DATA: 1WH:?OW7 IKS1
                                                                      CALI: i:Al.i:2«»l 0.1
    DASK II/E: ?:!*>
    P.Ii;:   3MOJ2.
                 I 0
    
                          128
                        123
                             135
                                    I?
    
    
                                                     4
                                                      I  T
                                                      LjL
                                                                                    255
                                                                                                     2?>2
                                                                                                                39360
    II/E      U50                     158                     2««)                     250                     3S9
    
    
      Figure A-5.  Electron impact  ionization mass  spectrum of  2,2',4,4'-tetrachlorobiphenyl.
    

    -------
         100.On
          59.8-
                         HAW s
                         8S/29/82 10:12:00.+ 1?.:RO
                         SAilPLC:  rut! UlXTUr.E—-SOLUTIOil (II  IWL IIU
                         conns.:  -iwo nnv IO-G 7oi;v .2ii.A ofl'j-riii 110-211-325-io/  -OJI-
                                 IS  1511 211 OT)
                  102
                                     ro
                                     III
                                                189
                                           171
                                       163
                            D.A1.A: '1(J:MI.:20V"/ QTffi
                            IJAI.l: <:AI.I:2CVM in
                                                                                      2.90
    n.Ast;
    II 1C:
                                                             2f
    lll^Ji|j.VL-r^r_|l)<
                                                                                                  I  ' I •' I"'
    
                                                                                                        350
          »/):     \W>               I'M               2W)              250              3«W
    
    
    Figure A-6.   Electron impact ionization mass  spectrum of 3,3'A,4'-tetrachlorobiphenyl-d5.
    

    -------
             100.0-1
              59.0-
                     99
                   J
                            HAS-; M'u
                            05/7:yi5:? 10:12:00 + 11:01
                            SAiiru;:  res MiXTunii—-S«I.UTIOII BI  sui. KM
                            CCSSUS.:  -I5'J9 CIIV 10-0 7«I:V .2IIA l)U:i I'jll 110-211-323-10/ "IJJI--COI.
                            EIIIIACCI'D (S 15B 211 Of)
    DATA:  1!>OIl:2c;Vlt/ (KIO'J
    CAM:  «JAI.F.2«IM 0.}
                              122
                                               163
                                                       |R1
    
                                          150
                                                                                   2H>
                                                                                 250
    BASK \1,'f.: 3>
    P.IO:   IIW'
                                                                                                               326
                      *" i
    
                   ^Jlk-r^-
                           300
                                                                                                                          3078
    Figure A-7.   Electron impact ionization mass  spectrum of 2,3',4,5',6-pentachlorobiphenyl.
    

    -------
    >
    
    I—1
    ~J
                  50.0-
                                IIACS su;cit:)in
                                e!i/28/«2  10:12:00 >  12:86
                                  irifi: rcu nixTunr—SOLUTION ni  IUL nu
                                      -I5b"0 CUV 10-5 70CV .2il\ 1)1)5-I'ill I
                                        (S I5D 211 01)
                             9
                  It'Ji
                                 128
                                                              2 0
                                                         2«0
    DATA: 19H1E20U07 U7?0 PASF: ll/t: %«
    CAM: CALI:20VI l)3 HIC: 323501.
    10-211-325- ID/ "051-COI."
    2
    2LM
    Iff?
    * 1 • 1 l"^^~l • 1 — ^
    250
    31
    ?e
    11
    325
    1 , .1 ,L.
    t—'r— *i— | • |-*"l "i i | i | W-*
    3tfO 350
    9
    
    1)
    1
                                                                                                                      25661.
               Figure A-8.   Electron impact ionization mass  spectrum of  2,2',3,3',6,6'-hexachlorobiphenyl.
    

    -------
    i
    i-1
    oo
    100.0-
                50.0
                   IIASS SPECTRUM
                   tfj/T.Vift IG-12-00 •>• I1-2I
                   SAurLE:  run iiixiui::-—-souirio!! ni  IIM. mi
                   COi!D:i.:  -I'jW BIV 10-0 701-V .2IIA ll.Vi-IMI 110-211-325-IO/ "OSJ-COI."
                   HKIAKCED (S 15B 211 OT)
    
    
                                                     1C?
                                                                                 DAM: isi.»iE2f.v»7 o««i
                                                                                 i;AJ.l : CAI.I-20UI flj
                                                                                              BA';E ii/i-: :;y-3
    
     ll^I        I
    
    UO.0-1
                50.0-
                     LI
                          2CT
                                          I2C
                                       170
    
                                    M
                           120
                                   271
                                                     11
                           200
                                       3»9
                iW
    32«
    31D
                                                                                       97
                                                                                                    220
                                                                                                     249
                                                                                                            3?'l
    3CO
    1 I ' ' ' ' '
    300
    400
                                                                                                                            ICOGt.
                                                                                                                        r   iceo
               Figure A-9.
                                Electron impact  ionization mass  spectrum of  2,2',3,4,4',5,6-heptachlorobiphenyl.
    
    

    -------
      I00.0-T
       50.0-
             95
                  I03
               v^-J
                  2?3
       50.0-
            271
               200
    SI'ECTRWI
    '«2 10:12:00 + |f>:H
    [•: rc» iiixiuni-: — SUUITIOII in mi. IH.J
    . : -1550 HIIV iO-O 70I-V .2IIA DilS-l'iSI II
    CED (S I5II 2II OT)
    IV
    
    125 132 «?2
    129 1
    129 HO ICQ 1
    3C
    
    323
    !f)0 3j| | |J t
    380 320 310 3
    PATA: l!HUI-2fitfl.«7 ll!)7l RAW: H/E: 170
    •JA.I.I: CAI.F.2CM 03 ItlC: .'JJCB11.
    0-2II-375-IO/ "Oil COI."
    4^
    2
    I9I>
    III OfVi
    l| T_|.r M |'[ l|l|_.» [ l| l|.( , |
    09 2C?0
    B
    R
    ,» -n In
    229 219 269
    393
    1
    -I, 1
    I
    Ll. . iL
    60 300 1»50 120 4i9
                                                                                                16G7:
                                                                                                IGG;
    Figure A-10.   Electron impact  ionization mass  spectrum of 2,2',3,3',4,4',5,6-octachlorobiphenyl.
    

    -------
         100.0-1
         50.0-
         100.0-
         50.0-
                        IIACS SiliCtilllil
                        05/20/1.7  lfi:42:OT '  10-. fj
                        SAi!n.i;i rcii iiixiur.;;- — SULHTIUII ui  tui. it.u
                        CMDS.: - TOO DIV IO-« 70RV .2IIA DU'i- riil 1 10-211-325- IO/ "051 -COI."
                        ElUIACCtD  (S I5lt 2tl OTI
                                                                  UG
                  DATA: l
    
    
    
    
    
    
    Jl
    
    
    
    
    ino
    
    
    i |
    
    
    190
    
    
    
    
    
    
    2
    luiiii 	 L)
    
    232
    
    
    
    
    1
    III f
    
    
    
    W
    
    Uit.. 	 I
                  IG9       120       140       100
                               322
                                                                                                                      17661.
               I   2?°   3?7
               il.-..il-ii.TTf-i
    
    320
                                                  3CO
    30!)
    4<
    *?s-.
    i-1
    
    4Wi 120 -110 1CO t«S
    Figure A-ll.   Electron  impact  ionization mass  spectrum  of  2,2',3,3',4,4',5,6,6'-nonachlorobiphenyl.
    

    -------
                                                                           I)\U:
                                                                           CAM: CAI.n2P.yi (13
    liAG'J SPECTRUM
    e5/?0/C2 10:12:00 +  17:13
    SAiirti-: rcn iii.vnir.r- -soiunnii in  nti. itu
    OHMS.: -1558 HIV 10-6 70I;V  .2IIA Dift-1511 110-211-325-IO/ "UJI-COL
            (S I5H 211 OT)
                                                211
    IBO.O-
    50.0-
    
    
    
    
    II/E
    198.0-
    
    
    50.0-
    
    170
    
    
    
    
    
    
    1
    
    96
    
    
    
    
    
    
    
    L ll.
    109
    
    
    
    
    32
    J
    
    
    17
    1?
    1 f
    1 ., J1J
    '
    
    
    
    
    ., ?ft ,
    
    
    I
    
    1
    I32
    JL|lLj
    •3
    I
    
    
    ISO
    
    
    
    35(
    
    i
    
    I ,
    
    
    0
    
    
    
    ir,e
    
    III
    
    
    
    
    379
    
    II,,
    
    
    
    
    
    
    
    
    
    IM
    L.
    
    *
    
    
    ,l.l
    
    
    
    
    
    
    
    
    
    f-
    
    2
    
    
    
    
    1 2?' 1
    1 t , • .[II. lilij 	 ,.1
    2
    
    
    
    1;
    393
    
    1
    1
    1
    B
    JU-,-.. , . ffi
    
    0
    
    
    
    
    ,. f? ?!?.
    2*6
    
    
    II. 29? .
    s~| 	 j •
    50 3f«
    
    
    100
    
    	 I.
    
    |
    I • 1
                                                                                                                    22016.
                                                                                                                 r  2281
              ll/'fi
                                    350
                                                        -150
    ses
    Figure A-12.   Electron  impact ionization mass  spectrum  of  2,2',3,3',4,4',5,5',6,6'-decachlorobiphenyl.
    

    -------
                            mt:0i.ii  HAW: 290  «nei;.coiir:0i.it IIASS: 2001
                 MUM) G TnTnAciii.ciiiHJmiBiYL. ISOKH: 0210
                 f:ti::D-ti n-iR.'.aii.o!:o--BiFi!i3M.. ISOIIKI: 11210
                 iAni;v/rn:.Ai:tA>/(AiiT./r.oi;.,\:iT.> (AV:
                                                                                                                       (i.OOO
    
                                                                                                                   % ST.DEY.-
                                                                                                                       O.OOD
                                                                                                                         i.o
    I
    N)
    ho
                      l.Ct;0
                      O.S90
                                                     -x-—5:—Jc	
                 IiATE
    5/1-1/32
    5/21/02
                                                                                        6/ 3/02
                Fie  re A-13    Response  factor plotted on a day-to-day basis  for the internal  standard,
    
                                      3,3' ,4,4'-tetrachlorobiphenyl-d6,  in Solution No. 1.
    

    -------
    I
    10
    RBI'OilSli    LIII:01.2i IIASS:  108  (i:i-:F.COIIP:QI. li
    DIP : 2-IIOHI111I5.0KO-DI PIIRIVI.. J SCIfEP, II 1
    P,tF:I)-G TtTPACIlLOnO-BirilB'.Y!.. ISOIIEB 112 U)
    (AI!L'A./»',rf .AI:J-A)/(AIIT../nb'F.AIIT. ) (AV:      '
         5.GOO
                                                                S:  2:"«)
                       1.0CO
                       3.C80
                       2.G90
                  DATK         5/1-1/82
                                                            X
    
                                                             t
                                                                                                                     si.i;;;1.'.-
                                                                                                                         0.431
                                                                                                                     X ST.riEY.-
                                                                                                                        12.171
                                                                                                                     Sl'il -1.9
                                                                                                               x   ;
                      Figure A-14.   Relative response'factor for 2-monochlorobiphenyl in Solution No. 1
    
                            calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a  day-to-day basis.
    

    -------
    I
    ro
                    ni-:sroi!5F.    1.10:01,1. MASS: 222  im-:i:.o»ir:oi.ii iiAr,5:  2001
                    aiP:3.3>-nic!ii.nr.o-nin:i:.:ivi..  ISOIIHR nil
                    r.GI::D-C TimiACIIIJinO-BMIIBIYL. ISOIini! 0210
                                      l!CT.AIIT.) (AV:      2.7;?2»
                                                             }
                         2.890
                         2.703
                         2.669
                         2.583
                         2.400
                         2.380
         2.200
    DATE         5/11/02
      sr.ntv.-
         0.206
      Tt. ST.DEY.-
                                                                                                                      51II - l.t
                                                                          5/21/02
    €./ 3/32
                   Figure A-15.   Relative  response  factor for 3,3'-dichlorobiphenyl in Solution No.  1
    
                         calculated versus  3,3',4,4'-tetrachlorobiphenyl-d^ on  a day-to-day  basis.
    

    -------
    RESPONSE    i.in:0i.i. MASS: 25G  
    uip:2.i.5-TRiaii.oi:o-niniEiiYi.. isotiui! n29
    REF:n-G TL:iR.\aii.ni;o-niriiiaiYi.. ISOIIEI: 11210
    (ARP..A/l!rr.Ai:i:A)/(AIIT./m:r.A!IT.) (AY:     2.085I
    
    
    
    
    
    2.2(in
    
    
    2. «00
    
    2.000
    
    
    
    
    ^>-
    i •
    Ul
    1.999
    
    1.883
    
    1 7IM4
    
    
    
    i
    ;
    
    
    '
    
    
    1
    
    
    
    
    
    
    X
    
    
    
    
    
    
    
    
    
    '
    f
    
    -
    
    
    
    
    
    
    
    
    
    
    
    
    >
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    X
    
    
    X
    h
    :
    >
    
    ST.D1-V.-
    0. 163
    ? ST.IEY.-
    ft. 138
    
    
    Si«l =1.8
    
    
    X
    1
    1
    t
    1
    *
    X
    
    
    
    
    
    :
    
                                                         5/24/82
    6/ 3/82
     Figure A-16.   Relative  response  factor for 2,4,5-trichlorobiphenyl in  Solution  No.  1
    
             calculated  versus 3,3',4,4'-tetrachlorobiphenyl-ds on  a day-to-day basis.
    

    -------
    >
    
    K>
                 Ri-srousr    rin:0i.5i MASS: 202  (ni-:F.conr:oi.ii MASS: 200
                 aiP:2.2' .i.r-Tp.ir.Aaii.or.o-BiriiuiiYi.. ISOHHR BIT
                 nEF:D-0 rEJUAaiLPRO-niPIIFJJVI.. ISOWR 1J2IO   •
                 (AR[-A/nB:.A!:i-AI/(Alir./!!l;l;.AIIT.) (AV:     1.0321
                      1.200
                      1.100
                      0.C90
                 DATK         5,'I'1/82
                                                          '•
    5/2-1/82
                                                                                                                  ST.DI-V.-
                                                                                                                     0.050
                                                                                                                  A ST.DEY.-
                                                                                                                     5.5%
                                                                                                                  S!« -1.9
    G/ 3/82
               Figure A-17.   Relative  response  factor for 2,2',4,4'-tetrachlorobiphenyl  in  Solution  No.  1
    
                          calculated versus  3,3',4,4'-tetrachlorobiphenyl-d6  on  a day-to-day  basis.
    

    -------
    nESPOIISK
                    I.IBrOI.C. MASS: 32G  (l!F.F.COIir:01. !•
                                      i.. isoni;n ni2i
                      ..
          (Ani-A/BF.P.Ai:nA»/(AIIT./riKF.AIIT.)  (AV:
              1.030
         0.899
    DATE         5/11/32
    \
    f
    )
    <
    -_l 	 _ — — -
    
    >
    (
                                                                                                      -ST.IiHY.-
                                                                                                          (1.052
                                                                                                      2 ST. COY.-
                                                                                                          >.477
                                                                                                       St'il
                                                             5/2-1/82
                                                                                               G/ 3/82
    Figure A-18.   Relative response  factor  for 2,3',4,5',6-pentachlorobiphenyl  in Solution No.  1
    
                calculated  versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis.
    

    -------
    I
    ro
    co
                              1.10:01.?! IIASS: 3co  u:i'F.co!ir:Gi.ii HASS:
                   air:2.2\3.3\6.6i-Mi:KAaiLcr.i)-BmiFjiYi..  ISOIIF.B« no
                   BF.F:0-G TE1KAClll.Oi:0-lliri!lilY!.. 1 SOI 11:1! (!2IO
                   (ARfiA/l!EF.Ar.!:A)/(AttT./IltH.AIIT.I  (AV;     0.6C5I
                        O.PO'.I
                        o./oo
                        O.C09
                        0.580
                   DATE         5/14/82
    5/2-1/82
                                                                                                                    ST. BEY.-
                                                                                                                       0.014
                                                                                                                    S ST.DEY.-
                                                                                                                       0.370
                                                                                                                    SHI - I.0
                                                                                                              X
    C/ 3/02
            Figure A-19.  Relative response factor  for  2,2',3,',6,6'-hexachlorobiphenyl in  Solution No.  1
    
                         calculated versus  3,3',4,4'-tetrachlorobiphenyl-d6 on  a day-to-day basis.
    

    -------
    ho
    vo
                       n    MD:OI.RI mss: 391 ii!nF.caiir:ni.ii  MASS:
                 air:2.2'.3.-i./i'.5.G-iici>T/,nii.or.o-BiriiEiivi.. isonnn n mi
                 BEF:B-c TKTnAUii.ono-iiiriieivi.. isonia; 11210
                 URFA/RHF./.RCAVtAriT./r.tF.AIIT.) (AV:     0.377>
    U.-1JW
    0. 110
    o.w -
    
    0.300
    0.380
    0.370
    0.360 -
    
    8.350
    0.310
    8.330
    o tin
    
    
    X
    ,f
    • i
    • i
    i
    i
    t
    f
    •
    1
    t
    I
    X
    
    
    
    )
    >
    
    
    
    
    
    
    
    
    
    c
    ST.M-V.-
    < 0.027
    S ST.IiEY.
    7.035
    5WI - 1.0
    
    X
    
    MX
    
    
    X
    X
    
                                                                      5/2-1/82
                                                                                                              G/ 3/82
           Figure A-20.   Relative response  factor for  2,2',3,4,4',5,6-heptachlorobiphenyl  in Solution No.  1
    
                        calculated  versus 3,3',4,4'-tetrachlorobiphenyl-d6 on  a day-to-day basis.
    

    -------
    .1
    OJ
    o
                 r.Ksro;isn    i.iD:oi.9i  IIASS: 130      .
                 CIIP:2.2l.3.3'.1.'T.5.6-l)lJTAf:MLOno-l)II'lli;ilYI..  ISOIIRR II  I95
                 REi::D--c TniiiAciiLor.o-Kiriii:>m. isounn 11210
                 (Ar.t:/./P,£F.AP.KA>/(AllT../P.\;i--.AllT.) (AY:     0.27fl)
                      0.30i>
                      o.:
                      0.269
                      0.250
                      o.;
                                                                                                                   sr.nr.v.-
                                                                                                                       0.021
                                                                                                                   z sr.ncv.-
                                                                                                                       7.072
                                                                                                                         1.0
                                                                                                                 X
    DATE
    5/14/82
                                                                       5/2-1/82
    C»/ 3/02
           Figure A-21.   Relative response factor  for  2,2',3,3',4,4',5,6-octachlorobiphenyl  in  Solution  No.  1
    
                          calculated versus  3,3',4,4'-tetrachlorobiphenyl-dg on a  day-to-day basis.
    

    -------
    I
    OJ
             DATE
    -c TnTfi.'.csitor.o-nminjiYL. ISOIIKR 11210
    /nni:.ARI:A)/(AMT./r.EF.AIir.) IAV: 0.2571
    0. 320
    U.3,0 -
    n.:•»
    
    
    
    
    X
    
    
    
    j
    
    
    
    
    
    
    
    
    
    
    
    
    i
    f j
    
    V
    >
    ST.MiV.-
    0.029
    ^ J: ST.|:EV.-
    tt.239
    SOU - 1.0
    
    
    
    
    
    ^•-
    
    
    
    X
    X
    
    <
    5/11/82
                                                          5/21/02
    6/
         Figure A-22.   Relative response factor for 2,2' ,3,3 ' ,4,4' ,5,6,6' -nonachlorobiphenyl in Solution No . ', 1
                      calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis.
    

    -------
    >
    
    ro
               RJISrOIISE    LlB:01.1t. 1IASS:  498  tEEF.OmMH. li MASS: 299)
               aiP!2.2>.3.3g.4.4'.5.5i.C.6i-DLCAail.OP,0-DII'lllillYL.  ISOHHIi D 209
               ni-F:D-t> TEU!/\aiLOBO-niPiiwri'L.  isoiiiiis 11210
               (AI',i;A/r,lil;.Ai!li'.>/(AHT./IlEl:.MIT.)  IAV:     0.223)
                    0.2/0
                    it. ?r.n
                    0.2SO
                    0.210
                    0.2JO
                    0.220
                    0.210
                    0.200
                    0. 100
               UATI:
                                 «
                                 *
                                                                                  ?
                                                                                  4
                                                                                                                 Sl.HSV.-
                                                                                                                     C.022
                                                                                                                 ,? sr.otv.-
                                                                                                                     0.077
                                                                                                                 StW "1.8
    5/H/82
    5/2-1/U2
                                                                                                               C/ 3/02
            Figure A-23.   Relative  response  factor  for 2,2',3,3',4,4',5,5',6,6'-decachlorobiphenyl in  Solution No.  1
    
                          calculated  versus 3,3' ,4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis.
    

    -------
                            APPENDIX B
    
    
    ANALYTICAL METHOD:  THE ANALYSIS OF BY-PRODUCT CHLORINATED
        BIPHENYLS IN COMMERCIAL PRODUCTS AND PRODUCT WASTES
                              B-l
    

    -------
                 THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS IN
                       COMMERCIAL PRODUCTS AND PRODUCT WASTES
    1.0   Scope and Application
    
          1.1   This is a gas  chromatographic/electron impact mass  spectrometric
                (GC/EIMS) method applicable to  the  determination of chlorinated
                biphenyls (PCBs) in commercial  products and product wastes.   The
                PCBs present may originate either as  synthetic by-products  or as
                contaminants derived from commercial  PCB products (e.g.,  Aroclors).
                The PCBs may be present as single isomers or complex mixtures and
                may include all 209 congeners  from  monochlorobiphenyl through
                decachlorobiphenyl listed in Table  1.
    
          1.2   The detection  and quantitation limits are dependent upon  the com-
                plexity of the sample matrix and the  ability of the analyst to
                remove interferents and properly maintain the analytical  system.
                The method accuracy and precision will be determined in future
                studies.
    
          1.3   This method is restricted to use by or under the supervision of
                analysts experienced in the use of  gas chromatography/mass  spec-
                trometry (GC/MS) and in the interpretation of gas chromatograms
                and mass spectra.  Prior to sample  analysis, each analyst must
                demonstrate the ability to generate acceptable results with this
                method by following the procedures  described in Section 14.2.
    
          1.4   The validity of the results depends on equivalent recovery  of the
                analyte and 13C PCBs.  If the  *3C PCBs are not thoroughly incor-
                porated in the matrix, the method is  not applicable.
    
          1.5   During the development and testing  of this method,  certain  analyti-
                cal parameters and equipment designs  were found to  affect the valid-
                ity of the analytical results.   Proper use of the method  requires
                that such parameters or designs must  be used as specified.   These
                items are identified in the text by the word "must."  Anyone wish-
                ing to deviate from the method  in areas so identified must  demon-
                strate that the deviation does  not  affect the validity of the data.
                Alternative test procedure approval must be obtained from the
                Agency.  An experienced analyst may make modifications to param-
                eters or equipment identified  by the  term "recommended."  Each
                time such modifications are made to the method, the analyst must
                repeat the procedure in Section 14.2.   In this case,  formal ap-
                proval is not  required, but the documented data from Section 14.2
                must be on file as part of the  overall quality assurance  program.
                                       B-2
    

    -------
                                    TABLE 1.   NUMBERING OF  PCB CONGENERS3
    No.
    
    1
    2
    3
    
    
    
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    
    
    
    1$
    17
    IS
    19
    20
    21
    22
    23
    24
    25
    26
    27
    23
    29
    33
    31
    32
    33
    34
    35
    36
    37
    33
    39
    
    
    
    40
    41
    42
    43
    44
    45
    46
    47
    48
    49
    SO
    SI
    
    Structure
    Honeehlorott
    -------
    2.0   Summary
          2.1   The process or product must be sampled such that the specimen col-
                lected for analysis is representative of the whole.   Statistically
                designed selection of the sampling position, time,  or discrete
                product units should be employed,   the sample must  be preserved .
                to prevent PCB loss prior to analysis.  Customary inventory stor-
                age may be adequate for products.   For intermediates, process sam-
                ples, and other non-product specimens, storage at 4°C with op-
                tional preservation at low pH is recommended.
    
          2.2   The sample is mechanically homogenized and subsampled if necessary.
                The sample is then spiked with four 13C PCB surrogates and the
                surrogates incorporated by further mechanical agitation.
    
          2.3   The surrogate-spiked sample is extracted and cleaned up at the
                discretion of the analyst.  Simple dilution or direct injection
                is permissible.  Possible extraction techniques include liquid-
                liquid partition, thermal desorption, and sorption onto resin
                columns followed by solvent desorption.  Cleanup techniques may
                include liquid-liquid partition, sulfuric acid cleanup, saponifi-
                cation, adsorption chromatography, gel permeation chromatography,
                or a combination of cleanup techniques.  The sample is diluted or
                concentrated to a final known volume for instrumental determina-
                tion.
    
          2.4   The PCB content of the sample extract is determined by capillary
                (preferred) or packed column gas chromatography/electron impact
                mass spectrometry (CGC/EIMS or PGC/EIMS) operated in the selected
                ion monitoring (SIM), full scan, or limited mass scan (LMS) mode.
    
          2.5   PCBs are identified^by comparison of their retention time and mass
                spectral intensity ratios to those in calibration standards.
    
          2.6   PCBs are quantitated against the response factors for a mixture
                of 11 PCB congeners, using the response of the 13C  surrogate to
                compensate for losses in workup and determination and instrument
                variability.
    
          2.7   The PCBs identified by the SIM technique may be confirmed by full
                scan CGC/EIMS, retention on alternate GC columns, other mass spec-
                trometric techniques, infrared spectrometry, or other techniques,
                provided that the sensitivity and selectivity of the technique are
                demonstrated to be comparable or superior to GC/EIMS.
    
          2.8   The analysis time is dependent on the extent of workup employed.
                The time required for instrumental analysis of a single sample,
                excluding data reduction and reporting, is about 30 to 45 rain.
    
          2.9   Appropriate quality control (QC) procedures are included to assess
                the performance of the analyst and estimate the quality of the re-
                sults.  These QC procedures include the demonstration of laboratory
                capability:  periodic analyst certification, the use of control
    
                                       B-4
    

    -------
                charts, and the analysis of blanks, replicates, and standard addi-
                tion samples.  A quality assurance (QA) plan must be developed for
                each laboratory.
    
          2.10  While several options are available throughout this method, the
                recommended procedure to be followed is:
    
                2.10.1   The sample is collected according to a scheme which per-
                         mits extrapolation of the sample data to the whole pro-
                         duct or product waste.
    
                2.10.2   The sample is preserved to prevent any loss of PCBs or
                         changes in matrix which may adversely affect recovery.
    
                2.10.3   The sample is mechanically homogenized and subsampled if
                         necessary.
    
                2.10.4   The sample is spiked with four 13C PCB surrogates
                         (4-chlorobiphenyl;  3,3',4,4'-tetrachlorobiphenyl;
                         2,2',3,3',5,5',6,6'-octachlorobiphenyl;  and decachloro-
                         biphenyl).
    
                2.10.5   Normally,  the sample is  extracted, although dilution may
                         also be used.
    
                2.10.6   The extract is cleaned up and  concentrated to an appro-
                         priate volume.
    
                2.10.7   An aliquot of the extract is analyzed by CGC/EIMS oper-
                         ated in the SIM mode. On-eolumn injections onto a 15-m
                         DB-5 capillary column, programmed (for toluene solutions)
                         from 110°  to 325°C at 10°'/min after a 2-min hold is used.
                         Helium at  45-cm/sec linear velocity is used as the carrier
                         gas.
    
                2.10.8   PCBs are identified by retention time and mass spectral
                         intensities.
    
                2.10.9   PCBs are quantitated against the response factors for a
                         mixture of 11 PCB congeners.
    
                2.10.10  The total  PCBs are obtained by summing the amounts for
                         each homolog found, and  the concentration is reported
                         as micrograms per gram.
    
    
    3.0   Interferences
    
          3.1   Method interferences may be caused by contaminants in solvents,
                reagents, glassware, and other sample processing hardware, leading
                to discrete artifacts and/or elevated baselines in the total ion
                current profiles.  All of these materials must be routinely demon-
                strated to be free  from interferences by the analysis of labora-
                tory reagent blanks as described  in Section 14.4.
    
                                       B-5
    

    -------
                3.1.1    Glassware must be scrupulously cleaned.   All glassware
                         is cleaned as soon as possible after use by rinsing with
                         the last solvent used.   This should be followed by deter-
                         gent washing with hot water and rinses with tap water and
                         reagent water.  The glassware should then be drained dry
                         and heated in a muffle furnace at 400°C for 15 to 30 min.
                         Some thermally stable materials,  such as PCBs, may not
                         be eliminated by this treatment.   Solvent rinses with
                         acetone and pesticide quality hexane may be substituted
                         for the muffle furnace heating.   Volumetric ware should
                         not be heated in a muffle furnace.   After it is dry and
                         cool, glassware should be sealed and stored in a clean
                         environment to prevent any accumulation of dust or other
                         contaminants.  It is stored inverted or capped with
                         aluminum foil.
    
                3.1.2    The use of high purity reagents and solvents helps to
                         minimize interference problems.   Purification of sol-
                         vents by distillation in all-glass  systems may be re-
                         quired.  All solvent lots must be checked for purity
                         prior to use.
    
          3.2   Matrix interferences may be caused by contaminants that are coex-
                tracted from the sample.  The extent of matrix interferences will
                vary considerably from source to source,  depending upon the nature
                and diversity of the sources of samples.
    4.0   Safety
          4.1   The toxicity or carcinogenicity of each reagent used in this
                method has not been precisely defined;  however, each chemical
                compound should be treated as a potential health hazard.   From
                this viewpoint, exposure to these chemicals must be reduced to
                the lowest possible level by whatever means available.   The la-
                boratory is responsible for maintaining a current awareness file
                of OSHA regulations regarding the safe  handling of the  chemicals
                specified in this method.  A reference  file of material data han-
                dling sheets should also be made available to all personnel in-
                volved in the chemical analysis.
    
          4.2   Polychlorinated biphenyls have been tentatively classified as
                known or suspected human or mammalian carcinogens.  Primary stan-
                dards of these toxic compounds should be prepared in a  hood.
                Personnel must wear protective equipment, including gloves and
                safety glasses.
    
                Congeners highly substituted at the meta and para positions and
                unsubstituted at the ortho positions are reported to be the most
                toxic.  Extreme caution should be taken when handling these com-
                pounds neat or in concentrated solutions.  This class includes
                3,3',4,4'-tetrachlorobiphenyl (both natural abundance and isotop-
                ically labeled).
    
                                       B-6
    

    -------
          4.3   Diethyl ether should be monitored regularly to determine the per-
                oxide content.   Under no circumstances should diethyl ether be
                used with a peroxide content in excess of 50 ppm,  as an explosion
                could result.  Peroxide test strips manufactured by EM Labora-
                tories (available from Scientific Products Company, Cat. No.
                P1126-8 and other suppliers) are recommended for this test.  Pro-
                cedures for removal of peroxides from diethyl ether are included
                in the instructions supplied with the peroxide test kit.
    
          4.4   Waste disposal must be in accordance with RCRA and applicable
                state rules.
    5.0   Apparatus and Materials
    
          5.1   Sampling containers - Amber glass bottles,  1-liter or other ap-
                propriate volume, fitted with screw caps lined with Teflon.
                Cleaned foil may be substituted for Teflon  if the sample is not
                corrosive.  If amber bottles are not available, samples should
                be protected from light using foil or a light-tight outer con-
                tainer.  The bottle must be washed, rinsed  with acetone or methy-
                lene chloride, and dried before use to minimize contamination.
    
          5.2   Glassware - All specifications are suggestions only.  Catalog
                numbers are included for illustration only.
    
                5.2.1    Volumetric flasks - Assorted sizes.
    
                5.2.2    Pipets - Assorted sizes, Mohr delivery.
    
                5.2.3    Micro syringes - 10.0 (Jl for packed column GC analysis,
                         1.0 (Jl for on-column GC analysis.
    
                5.2.4    Chromatographic column - Chromaflex, 400 mm long x 19 mm
                         ID (Kontes K-420540-9011 or equivalent).
    
                5.2.5    Gel permeation chromatograph - GPC Autoprep 1002 (An-
                         alytical Bio Chemistry Laboratories, Inc.) or equivalent.
    
                5.2.6    Kuderna-Danish Evaporative Concentrator Apparatus
    
                         5.2.6.1  Concentrator tube - 10 ml,  graduated (Kontes
                                  K-570050-1025 or equivalent).  Calibration must
                                  be checked.  Ground glass stopper size (519/22
                                  joint) is used to prevent evaporation of solvent.
    
                         5.2.6.2  Evaporative flask - 500 ml (Kontes K-57001-0500
                                  or equivalent).  Attached to concentrator tube
                                  with springs (Kontes K-662750-0012 or equiva-
                                  lent) .
    
                         5.2.6.3  Snyder column - Three ball macro (Kontes
                                  K-503000-0121 or equivalent).
                /
    
                                       B-7
    

    -------
    5.3   Balance - Analytical, capable of accurately weighing 0.0001 g.
    
    5.4   Gas chromatography/raass spectrometer system.
    
          5.4.1    Gas chromatograph - An analytical system complete with a
                   temperature programmable gas chromatograph and all re-
                   quired accessories including syringes,  analytical columns,
                   and gases.  The injection port must be  designed for on-
                   column injection when using capillary columns or packed
                   columns.  Other capillary injection techniques (split,
                   splitless, "Grob," etc.) may be used provided the per-
                   formance specifications stated in Section 7.1 are met.
    
          5.4.2    Capillary GC column - A 12-20 m long x  0.25 mm ID fused
                   silica column with a 0.25 |Jm thick DB-5 bonded silicone
                   liquid phase (J&W Scientific) is recommended.  Alternate
                   liquid phases may include OV-101, SP-2100, Apiezon L,
                   Dexsil 300, or other liquid phases which meet the perfor-
                   mance specifications stated in Section  7.1.
    
          5.4.3    Packed GC column - A 180 cm x 0.2 cm ID glass column
                   packed with 3% SP-2250 on 100/120 mesh  Supelcoport or
                   equivalent is recommended.  Other liquid phases which
                   meet the performance specifications stated in Section  7.1
                   may be substituted.
    
          5.4.4    Mass spectrometer - Must be capable of  scanning from 150
                   to 550 daltons every 1.5 sec or less, collecting at least
                   five spectra per chromatographic peak,  utilizing a 70-eV
                   (nominal) electron energy in the electron impact ioniza-
                   tion mode and producing a mass spectrum which meets all
                   the criteria in Table 2 when 50 ng of decafluorotriphenyl
                   phosphine [DFTPP, bis(perfluorophenyl)phenyl phosphine]
                   is injected through the GC inlet.  Any  GC-to-MS interface
                   that gives acceptable calibration points at 10 ng per
                   injection for each PCB isomer in the calibration stan-
                   dard and achieves all acceptable performance criteria
                   (Section 10) may be used.  Direct coupling of the fused
                   silica column to the MS is recommended.  Alternatively,
                   GC-to-MS interfaces constructed of all  glass or glass-
                   lined materials are recommended.  Glass can be deacti-
                   vated by silanizing with dichlorodimethylsilane.
    
          5.4.5    A computer system that allows the continuous acquisition
                   and storage on machine-readable media of all mass spectra
                   obtained throughout the duration of the chromatographic
                   program must be interfaced to the mass  spectrometer.
                   The data system must have the capability of integrating
                   the abundances of the selected ions between specified
                   limits and relating integrated abundances to concentra-
                   tions using the calibration procedures  described in this
                   method.  The computer must have software that allows
                                 B-8
    

    -------
       TABLE 2.  DFTPP KEY IONS AND ION ABUNDANCE CRITERIA	
    
    Mass                           Ion abundance criteria
    
    
    197                            Less than 1% of mass 198
    198                            100% relative abundance
    199                            5-9% of mass 198
    
    275                            10-30% of mass 198
    
    365                            Greater than 1% of mass 198
    
    441                            Present, but less than mass 443
    442                            Greater than 40% of mass 198
    443                            17-23% of mass 442
                             B-9
    

    -------
                         searching any GC/MS data file for ions of a specific mass
                         and plotting such ion abundances versus time or scan
                         number to yield an extracted ion current profile (EICP).
                         Software must also be available that allows integrating
                         the abundance in any EICP between specified time or scan
                         number limits.
    6.0   Reagents
          6.1   Solvents - All solvents must be pesticide residue analysis grade.
                New lots should be checked for purity by concentrating an aliquot
                by at least as much as is used in the procedure.
    
          6.2   Calibration standard congeners - Standards of the PCB congeners
                listed in Table 3 are available from Ultra Scientific, Hope,
                Rhode Island;  or Analabs, North Haven, Connecticut.
    
          6.3   Calibration standard stock solutions - Primary dilutions  of each
                of the individual PCBs listed in Table 3 are prepared by  weighing
                approximately 1-10 mg of material within 1% precision. The PCB
                is then dissolved and diluted to 1.0 ml with hexane.   The concen-
                tration is calculated in mg/ml.  The primary dilutions are stored
                at 4°C in screw-cap vials with Teflon cap liners.  The meniscus
                is marked on the vial wall to monitor solvent evaporation.  Pri-
                mary dilutions are stable indefinitely if the seals are maintained.
                The validity of primary and secondary dilutions must be monitored
                on a quarterly basis by analyzing four quality control check sam-
                ples (see Section 14.2).
    
          6.4   Working calibration standards - Working calibration standards are
                prepared that are similar in PCB composition and concentration to
                the samples by mixing and diluting the individual standard stock
                solutions.  Example calibration solutions are shown in Table 3.
                The mixture is diluted to volume with pesticide residue analysis
                quality hexane.  The concentration is calculated in ng/ml as the
                individual PCBs.  Dilutions are stored at 4°C in narrow-mouth,
                screw-cap vials with Teflon cap liners.  The meniscus is  marked
                on the vial wall to monitor solvent evaporation.   These secondary
                dilutions can be stored indefinitely if the seals are maintained.
                These solutions are designated "CSxxx," where the xxx is  used to
                encode the nominal concentration in ng/ml.
    
          6.5   Alternatively, certified stock solutions similar to those listed
                in Table 3 may be available from a supplier, in lieu of the pro-
                cedure described in Section 6.4.
    
          6.6   DFTPP standard - A 50-ng/Ml solution of DFTPP is prepared in ace-
                tone or another appropriate solvent.
    
          6.7   Surrogate standard stock solution - The four 13C-labeled  PCBs
                listed in Table 4 may be available from a supplier as a certi-
                fied solution.  This solution may be used as received or  diluted
                further.  These solutions are designated "SSxxx," where the xxx
                is used to encode the nominal concentration in [Jg/ml.
                                       B-10
    

    -------
     TABLE 3.   CONCENTRATIONS OF CONGENERS IN PCB CALIBRATION STANDARDS  (rig/ml)a
    
    Homolog
    1
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    4
    1
    4
    8
    10
    Congener
    no.
    1
    3
    7
    30
    50
    97
    143
    183
    202
    207
    209
    210 (IS)
    211 (RS)
    212 (RS)
    213 (RS)
    214 (RS)
    CS1000
    1,040
    1,000
    1,040
    1,040
    1,520
    1,740
    1,920
    2,600
    4,640
    5,060
    4,240
    255
    104
    257
    \ 407
    502
    CS100
    104
    100
    104
    104
    152
    174
    192
    260
    464
    506
    424
    255
    104
    257
    407
    502
    CS050
    52
    50
    52
    52
    76
    87
    96
    130
    232
    253
    212
    255
    104
    257
    407
    502
    CS010
    10
    10
    10
    10
    15
    17
    19
    26
    46
    51
    42
    255
    104
    257
    407
    502
    
    a  Concentrations given as examples only.
                                       B-ll
    

    -------
       TABLE 4.  COMPOSITION OF SURROGATE SPIKING SOLUTION (SS100) CONTAINING
                                  13C-LABELED PCBsa
    
    Congener
    no.
    211
    212
    213
    214
    
    
    Compound
    (!' ,2' ,3' ,4' ,5' ,6'-13C6)4-chlorobiphenyl
    (13Ci2)3,3',4,4' -tetrachlorobiphenyl
    (13C12)2,2! ,3,3' ,5,5* ,6,6'-octachlorobiphenyl
    (13C12)decachloirobiphenyl
    
    
    Concentration
    (Mg/ml)
    104
    257
    395
    502
    
    a  Concentrations given as examples only.
                                        B-12
    

    -------
          6.8   Internal standard solution - A solution of de-3,3*,4,4'-tetra-
                chlorobiphenyl is prepared at a nominal concentration of 1-10
                mg/ml in hexane.   The solution is further diluted to give a work-
                ing standard.
    
          6.9   Solution stability - The calibration standard,  surrogate, and
                DFTPP solutions should be checked frequently for stability.  These
                solutions should be replaced after 6 months, or sooner if compari-
                son with quality control check samples  indicates compound degrada-
                tion or concentration change.
    
          6.10  Quality control check samples will be supplied  by the Agency.
    
    
    7.0   Calibration
    
          7.1   The gas chromatograph must meet the minimum operating parameters
                shown in Tables 5 and 6, daily.  If all criteria are not met, the
                analyst must adjust conditions and repeat the test until all cri-
                teria are met.
    
          7.2   The mass spectrometer jmust meet the minimum operating parameters
                shown in Tables 2, 7, and 8, daily.  If all criteria are not met,
                the analyst must retune the spectrometer and repeat the test un-
                til all conditions are met.
    
          7.3   The PCB response factors (RF ) must be  determined using Equation
                7-1 for the analyte homologs?
    
                                An x M.
                          RF  = -r*	^                                Eq. 7-1
                            p   A.  x M                                  ^
                                 is    p
                where    RF  = response factor of a given PCB congener
    
                          A  = area of the characteristic ion for the PCB congener
                                 peak
    
                          M  = mass of PCB congener injected (nanograms)
    
                         A.  = area of the characteristic ion for the internal
                                 standard peak
    
                         M.  = mass of internal standard injected (nanograms)
                          IS
    
                Using the same conditions as for RF , the surrogate response
                factors (RF ) must be determined using Equation 7-2.
                           s
                                  A  x M.
                            RF  = ^	£5.                              Eq. 7-2
                              s   A.  x M                                H
                                   is    s
    
                where A  = area of the characteristic ion for the surrogate peak
                       S
    
                      M  = mass of surrogate injected (nanograms)
                       S
    
                Other terms are the same as defined in Equation 7-1.
                                       B-13
    

    -------
    TABLE 5.  OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM
            Parameter
          Recommended
        Tolerance
    Gas chromatograph
    Column
    
    Liquid phase
    
    Liquid phase thickness
    Carrier gas
    Carrier gas velocity
    Injector
    Injector, temperature
    Injection volume
    Initial column temperature
    Column temperature program
    Finnigan 9610
    15 m x 0.255 mm ID
    Fused silica
    DB-5 (J&W)
    0.25 |Jm
    Helium
    45 cm/sec
                   f+
    On-column (J&W)
    Optimum performance
    1.0 [Jlc
    70°C (2 min)d
    70°-325°C at 10°C/mine
    Other
    Other
    
    Other nonpolar
    or semipolar
    < 1 |Jm
    Hydrogen
    Optimum performance
    Other
    Optimum performance
    Other
    Other
    Other
    Separator
    Transfer line temperature
    Tailing factor
    Peak width1
    None
    280°C
    0.7-1.5
    7-10 sec
    Glass jet or othe
    Optimum*
    0.4-3
    < 15 sec
    
    a  Substitutions permitted with any common apparatus or technique provided
       performance criteria are met.
    b  Measured by injection of air or methane at 270°C oven temperature.
    c  For on-column injection, manufacturer's instructions should be followed
       regarding injection technique.
    d  With on-column injection, initial temperature equals boiling point of the
       solvent; in this instance, hexane.
    e  C12C11Q elutes at 270°C.  Programming above this temperature ensures a
       clean column and lower background on subsequent runs.
    f  Fused silica columns may be routed directly into the ion source to pre-
       vent separator discrimination and losses.
    g  High enough to elute all PCBs, but not high enough to degrade the column
       if routed through the transfer line.
    h  Tailing factor is width of front half of peak at 10% height divided by width
       of back half of peak at 10% height for single PCB congeners in solution CSxxx,
    i  Peak width at 10% height for a single PCB congener is CSxxx.
                                       B-14
    

    -------
     TABLE 6.   OPERATING PARAMETERS FOR PACKED COLUMN GAS CHROMATOGRAPHY SYSTEM
          Parameter
       Recommended
       Tolerance
    Gas chromatograph
    Column
    Finnigan 9610
    180 cm x 0.2 cm ID
    Other3
    Other
    Column packing
    
    
    Carrier gas
    
    Carrier gas flow rate
    
    Injector
    
    Injector temperature
    
    Injection volume
    
    Initial column temperature
    
    Column temperature program
    
    Separator
    
    Transfer line temperature
    
    Tailing factor
    
    Peak width
    glass
    
    31 SP-2250 on 100/
    120 mesh Supelcoport
    
    Helium
    
    30 ml/tnin
    
    On-column
    
    250°C
    
    1.0 Ml
    
    150°C, 4 min
    
    150°-260°C at 8°/min
    
    Glass jet
    
    280°C
    
    0.7-1.5
    
    10-20 sec
    Other nonpolar
    or semipolar
    
    Hydrogen
    
    Optimum performance
    
    Other
    
    ^  .    b
    Optimum
    
    ^ 5 Ml
    
    Other
    
    Other
    
    Other
           ft
    Optimum
    
    0.4-3
    
    < 30 sec
    a  Substitutions permitted if performance criteria are met,
    
    b  High enough to elute all PCBs.
    
    c  Tailing factor is width of front half of peak at 10% height divided by
       width of back half of peak at 10% height for single PCB congeners in solu-
       tion CSxxx.
    
    d  Peak width at 10% height for a  single PCB congener in CSxxx.
                                      B-15
    

    -------
       TABLE 7.  OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER SYSTEM
    
          Parameter                    Recommended                Tolerance
    Mass spectrometer
    Data system
    Scan range
    Scan time
    Resolution
    Ion source temperature
    Electron energy
    Trap current
    Multiplier voltage
    Preamplifier sensitivity
    Finnigan 4023
    Incos 2400
    95-550
    1 sec
    Unit
    280°C
    70 eV
    0.2 mA
    -1,600 V
    10"6 A/V
    Other3
    Other
    Other
    Otherb
    Optimum performance
    200°-300°C
    Optimum performance
    Optimum performance
    Optimum performance
    Set for desired
    working range
    
    a  Substitutions permitted if performance criteria are met.
    
    b  Greater than five data points over a GC peak is a minimum.
    
    c  Filaments should be shut off during solvent elution to improve instrument
       stability and prolong filament life, especially if no separator is used.
                                        B-16
    

    -------
     TABLE 8.  OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER SYSTEM
    
    Parameter
    Mass spectrometer
    Data system
    Scan range
    Scan mode
    Cycle time
    Resolution
    Ion source temperature
    Electron energy
    Emission current
    Filament current
    Multiplier
    Recommended
    Finnigan MAT 311A
    Incos 2400
    98-550
    Exponential
    1.2 sec
    1,000
    280°C
    70 eV
    1-2 mA
    Optimum
    -1,600 V
    Tolerance
    Other3
    Other
    Other
    Other
    Otherb
    > 500
    250°-300°C
    70 eV
    Optimum
    Optimum
    Optimum
    
    a  Substitutions permitted if performance criteria are met.
    
    b  Greater than five data points over a GC peak is a minimum.
    
    c  Filaments should be shut off during solvent elution to improve instrument
       stability and prolong filament life, especially if no separator is used.
                                       B-17
    

    -------
                If specific congeners are known to be present and if standards
                are available, selected RF values may be employed.   For general
                samples, solutions CSxxx and SSxxx or a mixture (Tables 3 and 4),
                with a similar level of internal standard (dg-3,31,4,4'-tetra-
                chlorobiphenyl) added, may be used as the response  factor solution.
                The PCB-surrogate pairs to be used in the RF calculation are listed
                in Table 9.
    
                Generally, only the primary ions of both the analyte and surrogate
                are used to determine the RF values.   If alternate  ions are to be
                used in the quantitation, the RF must be determined using that
                characteristic ion.
    
                The RF value must be determined in a manner to assure ±20% accu-
                racy and precision.  For instruments with good day-to-day preci-
                sion, a running mean (RF) based on seven values determined once
                each day may be appropriate.  Other options include, but are not
                limited to, triplicate determinations of a single concentration
                spaced throughout a day or determination of the RF  at three dif-
                ferent levels to establish a working curve.
    
                If replicate RF values differ by greater than ±10% RSD, the system
                performance should be monitored closely.  If the RSD is greater
                than ±20%, the data set must be considered invalid  and the RF re-
                determined before further analyses are done.
    
          7.4   If the GC/EIMS system has not been demonstrated to  yield a linear
                response or if the analyte concentrations are more  than two orders
                of magnitude different from those in the RF solution, a calibration
                curve must be prepared.  If the analyte and RF solution concentra-
                tions differ by more than one order of magnitude, a calibration
                curve should be prepared.  A calibration curve should be estab-
                lished with triplicate determinations at three or more concentra-
                tions bracketing the analyte levels.
    
          7.5   The relative retention time (RRT) windows for the 10 homologs and
                surrogates must be determined.  If all congeners are not available,
                a mixture of available congeners or an Aroclor mixture (e.g.,
                1016/1254/1260) may be used to estimate the windows.  The windows
                must be set wider than observed if all isomers are  not determined.
                Typical RRT windows for one column are listed in Table 10.  The
                windows may differ substantially if other GC parameters are used.
    
    
    8.0   Sample Collection, Handling, and Preservation
    
          8.1   Amber glass sample containers should have Teflon-lined screw caps.
                With noncorrosive samples, methylene chloride-washed aluminum foil
                liners may be substituted.  The volume and configuration are deter-
                mined by the amount of sample to be collected and its physical
                properties.  For dry powders, other containers such as heavy-walled
                polyethylene bags may be appropriate.
    
    
                                       B-18
    

    -------
                         TABLE 9.  PAIRINGS OF ANALYTE, CALIBRATION. AND SURROGATE COMPOUNDS
    
    Analyte
    Congener
    no.
    1
    2,3
    4-15
    16-39
    40-81
    82-127
    128-169
    170-193
    194-205
    206-208
    209
    Compound
    Calibration standard
    Congener
    no.
    2~Ci2H9Cl 1
    3- and 4-C12H9Cl 3
    C12H7C13
    C12H6C14
    C12H5C15
    C12H4C16
    
    C12H2Clg
    C12HClg
    
    30
    50
    97
    143
    183
    202
    207
    209
    2
    4
    2,4
    2,4,
    2,2'
    2,2'
    2,2'
    2,2'
    2,2'
    2,2'
    p r>
    L12L
    
    6
    ,4
    ,3
    ,3
    ,3
    ,3
    ,3
    li
    Compound
    
    
    ,6
    ',4,
    ,4,5
    ',4,
    ,3',
    3'
    , j ,
    0
    
    
    
    5
    ,6'
    4', 5', 6
    5, 5', 6, 6'
    4, 4', 5, 6, 6'
    
    Surrogate
    Congener
    no.
    211
    211
    211
    212
    212
    212
    212
    213
    213
    213
    214
    Compound
    13C6-4
    J3C6-4
    13C12-3,3'
    13C12-3,3'
    13C12-3,3'
    13C12-3,3'
    13C12-2,2'
    13C12-2,2'
    13C12-2,2'
    13Ci2Cl10
    
    ,4
    ,4
    ,4
    ,4
    ,3
    ,3
    ,3
    
    
    ,4'
    ,4'
    ,4'
    ,4'
    ,3',5,5'
    ,3', 5, 5'
    3 ' 55*
    
    
    
    
    
    
    ,6,6'
    ,6,6'
    ,6,6'
    
    
    a  Ballschmiter numbering system, see Table 1.
    

    -------
           TABLE 10.  RELATIVE RETENTION TIME (RRT) RANGES OF PCB HOMOLOGS
                       VERSUS d6-3.3',4,4'-TETRACHLOROBIPHENYL
    
    PCB
    homolog
    Monochloro
    
    Dichloro
    Trichloro
    Tetrachloro
    Pentachloro
    Hexachloro
    Heptachloro
    Octachloro
    Nonachloro
    Decachloro
    No. of
    isomers
    measured
    3
    
    10
    9
    16
    12
    13
    4
    6
    3
    1
    Observed range
    of RRTsa
    0.40-0.50
    
    0.52-0.69
    0.62-0.79
    0.72-1.01
    0.82-1.08
    0.93-1.20
    1.09-1.30
    1.19-1.36
    1.31-1.42
    1.44-1.45
    Calibration
    Congener
    no.
    1
    3
    7
    30
    50
    97
    143
    183
    202
    207
    209
    solution
    Observed
    RRTa
    0.43
    0.50
    0.58
    0.65
    0.75
    0.98
    1.05
    1.15
    1.19
    1.33
    1.44
    Projected
    range of
    RRTs
    0.35-0.55
    
    0.35-0.80
    0.35-1.10
    0.55-1.05
    0.80-1.10
    0.90-1.25
    1.05-1.35
    1.10-1.50
    1.25-1.50
    1.35-1.50
    
    a  The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were
       measured versus 3,3',4,4'-tetrachlorobiphenyl-dg (internal standard) using
       a 15-m J&W DB-5 fused silica column with a temperature program of 110°C
       for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an on-
       column injector.  A Finnigan 4023 Incos quadrupole mass spectrometer oper-
       ating with a scan range of 95-550 daltons was used to detect each PCB
       congener.
    
    b  The projected relative retention windows account for overlap of eluting
       homologs and take into consideration differences in operating systems
       and lack of all possible 209 PCB congeners.
                                         B-20
    

    -------
    8.2   Sample bottle preparation
    
          8.2.1    All sample containers and caps should be washed in deter-
                   gent solution, rinsed with tap water, and then with dis-
                   tilled water.  The bottles and caps are allowed to drain
                   dry in a contaminant-free area.  Then the caps are rinsed
                   with pesticide grade hexane and allowed to air dry.
    
          8.2.2    Sample bottles are heated to 400°C for 15 to 20 min or
                   rinsed with pesticide grade acetone or hexane and allowed
                   to air dry.
    
          8.2.3    The clean bottles are stored inverted or sealed until use.
    
    8.3   Sample collection
    
          8.3.1    The primary consideration in sample collection is that
                   the sample collected be representative of the whole.
                   Therefore, sampling plans or protocols for each individ-
                   ual producer's situation will have to be developed.  The
                   recommendations presented here describe general situa-
                   tions.  The number of replicates and sampling frequency
                   also must be planned prior to sampling.
    
          8.3.2    Discrete product units - If the product is small enough
                   that one or more discrete units would be used as the an-
                   alytical sample, a statistically random sampling approach
                   is recommended.
    
          8.3.3    Liquids or free-flowing solids - If possible, the source
                   is mixed thoroughly before collecting the sample.  If
                   mixing is impractical, the sample should be collected
                   from a representative area of the source.  If the liquid
                   is flowing through an enclosed system, sampling through
                   a valve should be randomly timed.
    
          8.3.4    Solids - Larger bulk solids which must be subsampled to
                   get a reasonably sized analytical sample must be treated
                   on a case-by-case basis.  A representative sample should
                   be obtained by designing a sampling location selection
                   scheme such that all parts of the whole have a finite,
                   known probability of inclusion.  Based on such a scheme,
                   the PCB content of the sample can be used to extrapolate
                   to the content of the whole.
    
    8.4   Sample preservation - Product samples should be stored as the bulk
          or packaged product inventory would be stored, or in a cool, dry,
          dark area.  Intermediates, process samples, or other non-product
          specimens should be stored at 4°C.  If there is a possibility of
          microbial degradation, addition of H2S04 during collection to a
          pH < 2 is recommended.  A test strip is used to monitor pH.  Stor-
          age times in excess of 4 weeks are not recommended.
                                 B-21
    

    -------
                If residual chlorine is present in the sample,  it should be
                quenched with sodium thiosulfate.   EPA Methods  330.4 and 330.5
                may be used to measure the residual chlorine.1   Field test kits
                are available for this purpose.
    
    9.0   Sample Preparation
    
          Since a wide variety of matrices may be  subjected to  analysis by this
          method, the extraction/cleanup procedure cannot be specified.  This
          section describes general guidelines for subsampling, addition of 13C
          surrogates, dilution, extraction, cleanup, extract concentration, and
          other sample preparation procedures.
    
          9.1   Sample homogenization and subsarapling - The sample is homogenized
                by shaking, blending, shredding,  crushing, or other appropriate
                mechanical technique.  A representative subsample of 100 g or other
                known mass is then taken.  The sample size is dependent upon the
                anticipated PCB levels and difficulty of the subsequent extraction/
                cleanup steps.
    
                Note:  The precision of the mass  determination  at this step will
                be reflected in the overall method precision.  Therefore, an an-
                alytical balance must be used to  assure that the weight is accu-
                rate to ±1% or better.
    
          9.2   Surrogate addition - An appropriate volume of surrogate solution
                SSxxx is pipetted into the sample.  The final concentration of the
                surrogates must be in the working  range of the  calibration and
                well above the matrix background.   The surrogates are thoroughly
                incorporated by further mechanical agitation.  For nonviscous
                liquids, shaking for 30 sec should be sufficient.  For viscous
                liquids or free-flowing solids, 10-min tumbling is recommended.
                In cases where inadequate incorporation may be  expected, such as
                solids, overnight equilibration with agitation  is recommended.
    
                Note:  The volume measurement of  the spiking solution is critical
                to the overall method precision.   The analyst must exercise cau-
                tion that the volume is known to  ±J% or better.  Where necessary,
                calibration of the pipet is recommended.
    
          9.3   Sample preparation (extraction/cleanup) - After addition of the
                surrogates, the sample is further  treated at the discretion of
                the analyst, provided that the GC/EIMS response of the four sur-
                rogates meets the criteria listed  in Section 7.0.  The literature
                pertaining to these techniques has been reviewed.2  Several pos-
                sible techniques are presented below for guidance only.  The ap-
                plicability of any of these techniques to a specific sample ma-
                trix must be determined by the precision and accuracy of the 13C
                PCB surrogate recoveries, as discussed in Section 14.2.
                                       B-22
    

    -------
    9.3.1    Extraction1
    
             9.3.1.1  Dilution - In some cases,  where the PCB concen-
                      tration is high,  a simple  volumetric dilution
                      with an appropriate solvent may be sufficient
                      sample preparation.
    
             9.3.1.2  Direct injection - If sample viscosity permits,
                      direct injection with no dilution is permissible.
    
             9.3.1.3  Liquid-liquid extraction - If the matrix is
                      aqueous (or another solvent in which PCBs are
                      only slightly soluble),  a  liquid-liquid parti-
                      tion may be effective.  The solvent, number of
                      extractions, solvent-to-sample ratio, and other
                      parameters are chosen at the analyst's discretion.
    
             9.3.1.4  Sorbent column extraction - PCBs may be isolated
                      from free-flowing liquids  onto sorbent columns.
                      The selection of sorbent (XAD, Porapak, carbon-
                      polyurethane foam, etc.) will depend on the na-
                      ture of the matrix.  The available methods have
                      been reviewed.2
    
             9.3.1.5  Thermal desorption - If the matrix is nonvol-
                      atile, thermal desorption of the PCBs onto a
                      sorbent column, filter,  or cold trap may be an
                      effective extraction/cleanup method.
    
    9.3.2    Cleanup - Several tested cleanup techniques are described
             below.  All but the base cleanup (9.3.2.8) were previously
             validated for PCBs in transformer fluids.3  Depending
             upon the complexity of the sample,  one or more of the
             techniques may be required to fractionate the PCBs from
             interferences.  For most cleanups a concentrated (1-5 ml)
             extract should be used.
    
             9-3.2.1  Acid cleanup
    
                      9.3.2.1.1  Place 5 ml of concentrated sulfuric
                                 acid into a 40-ml narrow-mouth screw-
                                 cap bottle.  Add the sample extract.
                                 Seal the bottle with a Teflon-lined
                                 screw cap and shake for 1 min.
    
                      9.3.2.1.2  Allow the phases to separate, transfer
                                 the sample (upper phase) with three
                                 rinses of 1-2 ml solvent to a clean
                                 container and concentrate to an ap-
                                 propriate volume.
                           B-23
    

    -------
             9.3.2.1.3  Analyze as described in Section 10.0.
    
             9.3.2.1.4  If the sample is highly contaminated,
                        a second or third acid cleanup may
                        be employed.
    
    9.3.2.2  Florisil column cleanup
    
             9.3.2.2.1  Variations among batches of Florisil
                        (PR grade or equivalent) may affect
                        the elution volume of the various
                        PCBs.   For this reason, the volume
                        of solvent required to completely
                        elute  all PCBs must be verified by
                        the analyst.   The weight of Florisil
                        can then be adjusted accordingly.
    
             9.3.2.2.2  Place  a 20-g charge of Florisil,
                        activated overnight at 130°C, into a
                        Chromaflex column.  Settle the Flor-
                        isil by tapping the column.  Add about
                        1 cm of anhydrous sodium sulfate to
                        the top of the Florisil.  Pre-elute
                        the column with 70-80 ml of hexane.
                        Just before the exposure of the sodium
                        sulfate layer to air, stop the flow.
                        Discard the eluate.
    
             9.3.2.2.3  Add the sample extract to the column.
    
             9.3.2.2.4  Carefully wash down the inner wall
                        of the column with 5 ml of hexane.
    
             9.3.2.2.5  Add 200 ml of 6% ethyl ether/hexane
                        and set the flow to about 5 ml/min.
    
             9.3.2.2.6  Collect 200 ml of eluate in a Kuderna-
                        Danish flask.  All the PCBs should be
                        in this fraction.  Concentrate to an
                        appropriate volume.
    
             9.3.2.2.7  Analyze the sample as described in
                        Section 10.0.
    
    9.3.2.3  Alumina column cleanup
    
             9.3.2.3.1  Adjust the activity of the alumina
                        (Fisher A450 or equivalent) by heat-
                        ing to 200°C for 2 to 4 hr.  When
                        cool,  add 3% water (wt:wt) and mix
                        until  uniform.  Store in a tightly
                        sealed bottle.  Allow the deactivated
                        alumina to equilibrate at least 1/2 hr
                        before use.  Reactivate weekly.
                  B-24
    

    -------
             9.3.2.3.2  Variations  between batches  of alumina
                        may affect  the  elution volume of the
                        various  PCBs.   For this reason,  the
                        volume of solvent  required  to com-
                        pletely  elute all  of  the PCBs must
                        be  verified by  the analyst.   The
                        weight of alumina  can then  be ad-
                        justed accordingly.
    
             9.3.2.3.3  Place a  50-g charge of alumina into
                        a Chromaflex column.   Settle  the
                        alumina  by  tapping.   Add about 1 cm
                        of  anhydrous sodium sulfate.   Pre-
                        elute the column with 70-80 ml of
                        hexane.  Just before  exposure of the
                        sodium sulfate  layer  to air,  stop
                        the flow.   Discard the eluate.
    
             9.3.2.3.4  Add the  sample  extract to the column.
    
             9.3.2.3.5  Carefully wash  down the inner wall
                        of  the column with 5  ml of  hexane.
    
             9.3.2.3.6  Add 295  ml  of hexane  to the column.
    
             9.3.2.3.7  Discard  the first  50  ml.
    
             9.3.2.3.8  Collect  250 ml  of  the hexane  in a
                        Kuderna-Danish  flask.  All  of the
                        PCBs should be  in  this fraction.
                        Concentrate to  an  appropriate volume.
    
             9.3.2.3.9  Analyze  the sample as described in
                        Section  10.0.
    
    9.3.2.4  Silica gel column cleanup
    
             9.3.2.4.1  Activate silica gel (Davison  Grade
                        950 or equivalent) at 135°C overnight.
    
             9.3.2.4.2  Variations  between batches  of silica
                        gel may  affect  the elution  volume of
                        the various PCBs.   For this reason,
                        the volume  of solvent required to
                        completely  elute all  of the PCBs must
                        be  verified by  the analyst.  The
                        weight of silica gel  can then be ad-
                        justed accordingly.
                  B-25
    

    -------
             9.3.2.4.3  Place a 25-g charge of activated
                        silica gel into a Chromaflex column.
                        Settle the silica gel by tapping the
                        column.  Add about 1 cm of anhydrous
                        sodium sulfate to the top of the
                        silica gel.
    
             9.3.2.4.4  Pre-elute the column with 70-80 ml
                        of hexane.  Discard the eluate.  Just
                        before exposing the sodium sulfate
                        layer to air, stop the flow.
    
             9.3.2.4.5  Add the sample extract to the column.
    
             9.3.2.4.6  Wash down the inner wall of the column
                        with 5 ml of hexane.
    
             9.3.2.4.7  Elute the PCBs with 195 ml of 10%
                        diethyl ether in hexane (v:v).
    
             9.3.2.4.8  Collect 200  ml of the eluate in a
                        Kuderna-Danish flask.  All of the
                        PCBs should  be in this fraction.
                        Concentrate  to an appropriate volume.
    
             9.3.2.4.9  Analyze the  sample as described in
                        Section 10.0.
    
    9.3.2.5  Gel permeation cleanup
    
             9.3.2.5.1  Set up and calibrate the gel permea-
                        tion chromatograph with an SX-3 column
                        according to the Autoprep instruction
                        manual.  Use 15% methylene chloride
                        in cyclohexane (v:v) as the mobile
                        phase.
    
             9.3.2.5.2  Inject 5.0 ml of the sample extract
                        into the instrument.  Collect the
                        fraction containing the PCBs (see
                        Autoprep operator's manual) in a
                        Kuderna-Danish flask equipped with
                        a 10-ml ampul.
    
             9.3.2.5.3  Concentrate  the PCB fraction to an
                        appropriate  volume.
    
             9.3.2.5.4  Analyze the  sample as described in
                        Section 10.0.
                  B-26
    

    -------
    9.3.2.6  Acetonitrile partition
    
             9.3.2.6.1  Place the sample extract into a 125-ml
                        separately funnel with enough hexane
                        to bring the final volume to 15 ml.
                        Extract the sample four times by shak-
                        ing vigorously for 1 min with 30-ml
                        portions of hexane-saturated acetoni-
                        trile.
    
             9.3.2.6.2  Combine and transfer the acetonitrile
                        phases to a 1-liter separatory funnel
                        and add 650 ml of distilled water
                        and 40 ml of saturated sodium chloride
                        solution.  Mix thoroughly for about
                        30 sec.  Extract with two 100-ml por-
                        tions of hexane by vigorously shaking
                        about 15 sec.
    
             9.3.2.6.3  Combine the hexane extracts in a
                        1-liter separatory funnel and wash
                        with two 100-ml portions of distilled
                        water.  Discard the water layer and
                        pour the hexane layer through an 8-10
                        cm anhydrous sodium sulfate column
                        into a 500-ml Kuderna-Danish flask
                        equipped with a 10-ml ampul.  Rinse
                        the separatory funnel and column with
                        three 10-ml portions of hexane.
    
             9.3.2.6.4  Concentrate the extracts to an ap-
                        propriate volume.
    
             9.3.2.6.5  Analyze as described in Section 10.0.
    
    9.3.2.7  Florisil slurry cleanup
    
             9.3.2.7.1  Place the sample extract into a 20-ml
                        narrow-mouth screw-cap container.
                        Add 0.25 g of Florisil (PR grade or
                        equivalent).  Seal with a Teflon-lined
                        screw cap and shake for 1 min.
    
             9.3.2.7.2  Allow the Florisil to settle; then
                        decant the treated solution into a
                        second container with rinsing.  Con-
                        centrate the sample to an appropriate
                        volume.  Analyze as described in Sec-
                        tion 10.0.
                  B-27
    

    -------
    9.3.2.8  Base cleanup4
             9.3.2.8.1  Quantitatively transfer the concen-
                        trated extract to a 125-ral extraction
                        flask with the aid of several small
                        portions of solvent.
    
             9.3.2.8.2  Evaporate the extract just to dryness
                        with a gentle stream of dry filtered
                        nitrogen, and add 25 ml of 2.5% alco-
                        holic KOH.
    
             9.3.2.8.3  Add a boiling chip, put a water con-
                        denser in place,  and allow the solu-
                        tion to reflux on a hot plate for 45
                        min.
    
             9.8.2.8.4  After cooling, transfer the solution
                        to a 250-ml separatory funnel with
                        25 ml of distilled water.
    
             9.3.2.8.5  Rinse the extraction flask with 25
                        ml of hexane and  add it to the
                        separatory funnel.
    
             9.3.2.8.6  Stopper the separatory funnel and
                        shake vigorously  for at least 1 min.
                        Allow the layers  to separate, and
                        transfer the lower aqueous phase to
                        a second separatory funnel.
    
             9.3.2.8.7  Extract the saponification solution
                        with a second 25-ml portion of hexane.
                        After the layers  have separated, add
                        the first hexane  extract to the sec-
                        ond separatory funnel and transfer
                        the aqueous alcohol layer to the
                        original separatory funnel.
    
             9.3.2.8.8  Repeat the extraction with a third
                        25-ml portion of  hexane.  Discard
                        the saponification solution, and com-
                        bine the hexane extracts.
    
             9.3.2.8.9  Concentrate the hexane layer to an
                        appropriate volume, and analyze as
                        described in Section 10.0.
                  B-28
    

    -------
    10.0  Gas Chromatographic/Electrop Impact Mass Spectrometric Determination
    
          10.1  Internal standard addition - An appropriate volume of the internal
                standard solution is pipetted into the sample.   The final concen-
                tration of the internal standard must be in the working range of
                the calibration and well above the matrix background.  The inter-
                nal standard is thoroughly incorporated by mechanical agitation.
    
                Note:  The volumetric measurement of the internal standard solu-
                tion is critical to the overall method precision.  The analyst
                must exercise caution that the volume is known to be ±1% or better.
                Where necessary, calibration of the pipet is recommended.
    
          10.2  Tables 2, and 5 through 8 summarize the recommended operating con-
                ditions for analysis.  Figure 1 presents an example of a chromato-
                gram.
    
          10.3  While the highest available chromatographic resolution is not a
                necessary objective of this protocol, good chromatographic per-
                formance is recommended.  With the high resolution of CGC, the
                probability that the chromatographic peaks consist of single com-
                pounds is higher than with PGC.  Thus, qualitative and quantita-
                tive data reduction should be more reliable.
    
          10.4  After performance of the system has been certified for the day
                and all instrument conditions set according to Tables 2, and 5
                through 8, inject an aliquot of the sample onto the GC column.
                If the response for any ion, including surrogates and internal
                standards, exceeds the working range of the system, dilute the
                sample and reanalyze.  If the responses of surrogates, internal
                standards, or analytes are below the working range, recheck the
                system performance.  If necessary, concentrate the sample and re-
                analyze .
                                    *
          10.5  Record all data on a digital storage device (magnetic disk, tape,
                etc.) for qualitative and quantitative data reduction as discussed
                below.
    11.0  Qualitative Identification
    
          11.1  Selected ion monitoring (SIM) or limited mass scan (IMS) data -
                The identification of a compound as a given PCB homolog requires
                that two criteria be met:
    
                11.1.1   (1) The peak must elute within the retention time window
                         set for that homolog (Section 7.5); and (2) the ratio of
                         two ions obtained by SIM (Table 11) or by IMS (Table 12)
                         must match the natural ratio within ±20%.  The analyst
                         must search the higher mass windows, in particular M+70,
                         to prevent misidentification of a PCB fragment ion cluster
                         as the parent.
    
    
                                       B-29
    

    -------
               100.(I-,
                 nic_
    w
    i
    UJ
    o
    c
    O)
                              CX
                              •H
                              O
    
                              o
                              u
                              o
                              d
                              o
                             CM
                                   o
                                   CO
                                 c  c
                                 01  01
                                 rC f.
                                 cx  ex
                                 •H -H
    
                                 O  O OJ
                                 t-t  r-l 4-1
                                 O  O n)
                                 rH rH 60
                                 ,C ^ O
                                 U  0 VJ
                                 O  O M
                                 C  C p
                                 o  o to
                                 & & ^
                                  C7'J
               c
               Ol
    
               ex
              •H
              .£>
               O
               r-l
               O
              rH
              ,£
               O
              •H
              P
    C
    01
    
    cx
    •H
    .a
    o
    t-i
    o
    M
    H
                                        T   T
                                  _.L_JI	Lm.
                           c
                           O)
                                                    p.
     o
     r-l
     O
     rH
     ^2
     O
     cfl
     t-j
     4J
     Ol
     H
     I
     vO
     CM
    
     CM
    
    
    
    
    U7-I
                               	1 J^.
                            10:03 ,
    -a
    U)
    
    r-t
    n)
    c
    r-l
    QJ
    4-J
    c
    •rH
    " —
    \o
    •a
    rH
    c
    O)
    cx
    •H
    ,0
    O
    r-^lf"
    rH
    0
    Cfl
    r-4
    4J
    01
    H
    1
    -
    rH -
    rC *•
    CX CO
    •H
    ,£> CO
    O
    ^
    O
    rH
    f.
    0
    CO
    4-1
    C
    01
    PH
    1
    tn
    "*-
    co
    ™ (
    1
    
    
    01
    4-1
    Cfl
    60
    O
    1-1
    r-l
    (fl
    v — '
    u
    co
    rH
    1
    rH
    C
    01
    ex
    •H
    c\
    O
    V't*
    rH
    U
    CO rH
    r-l >*»
    4-j n
    11 0)
    H ,C
    1 CX
    •H
    -.
    •.
    m
    r.
    —
    ~^
    
    
                                                      O
                                                      M
                                                      O
                                                      O
                                                      CO
                                                      O
                                                      Ol
                                                      Q
                                                                           O
                                                                          co
                                                     C
                                                     Ol
    
                                                     cx ^.
                                                     •H Ol
                                                     ,0 4J
                                                     O CO
                                                     r-l 60
                                                     O O
                                                     rH t-l
                                                     f. r-l
                                                     O 3
                                                     cfl W
                                                     O '"S
                                                     O)
                                                     O
                                                                                                 2.3:20
                                                                                                          SCA'l
                                                                                                          Tlllii
           Figure  1.   Capillary gas  chromatography/electron impact  ionization mass  spectrometry  (CGC/EIMS)
    
             chromatogram or the calibration standard  solution required for quantitation of PCBs  by homolog.
    
             This  chromatogram includes  PCBs representative of each homolog, three  carbon-13 labeled surrogates,
    
             and the  deuterated internal standard.  The concentration  of a]1 components and the CGC/EIMS
    
             parameters  are presented  in Tables 3, 4,  5,  and 7.
    

    -------
                     TABLE 11.  CHARACTERISTIC SIM IONS FOR PCBs
                                              Ion (relative intensity)
    Homolog      ..                 Primary           Secondary          Tertiary
    C12H9C1 188 (100)
    C12H8C12 222 (100)
    Ci2H7Cl3 256 (100)
    C12H6C14 292 (100)
    Ci2H5Cl5 326 (100)
    Ci2H4Cl6 360 (100)
    C12H3C17 394 (100)
    C12H2C18 430 (100)
    Ci2HCl9 464 (100)
    C12Clio 498 (100)
    190 (33)
    224 (66)
    258 (99)
    290 (76)
    328 (66)
    362 (82)
    396 (98)
    432 (66)
    466 (76)
    500 (87)
    -
    226 (11)
    260 (33)
    294 (49)
    324 (61)
    364 (36)
    398 (54)
    428 (87)
    462. (76)
    496 (68)
    
    Source:  Rote, J. W., and W. J. Morris, "Use of Isotopic Abundance Ratios in
             Identification of Polychlorinated Biphenyls by Mass Spectrometry,"
             J. Assoc. Offic. Anal. Chero.. 56(1), 188-199 (1973).
                                        B-31
    

    -------
                TABLE 12.  LIMITED MASS SCANNING (LMS) RANGES FOR PCBs
    
    Compound
    v< 1 ^ITQ \j J_ 1
    LJ 1 Oil O \j 1 O
    C12H7C13
    C12H6C13
    Ci2H5Cl5
    C12H4C16
    C12H3C17
    C12H2C18
    C12HC19
    C12C110
    C12D6C14
    13C612C6H9C1
    13C12H6C14
    13C12H2C18
    13C12Clio
    Mass range (m/z)
    186-190
    220-226
    254-260
    288-294
    322-328
    356-364
    386-400
    426-434
    460-468
    494-504
    294-300
    192-196
    300-306
    438-446
    506-516
    
    a  Adapted from Tindall, G. W.,  and P. E. Wininger, "Gas Chromatography-Mass
       Spectrometry Method for Identifying and Determining Polychlorinated Bi-
       phenyls," J. Chromatogr.,  196, 109-119 (1980).
                                         B-32
    

    -------
                11.1.2    If one  or  the  other of these  criteria  is  not met,  inter-
                     . -   ferences may have  affected  the  results, and  a reanalysis
                         using full scan EIMS conditions is  recommended.
    
          11.2  Full scan data
    
                11.2.1    The peak must  elute within  the  retention  time windows
                         set for that homolog (as  described  in  Section 7.5).
    
                11.2.2    The unknown spectrum must match that of an authentic PCB.
                         The intensity  of the three  largest  ions in the molecular
                         cluster (two largest for  monochlorobiphenyls) must match
                         the natural ratio  within  ±20%.   Fragment  clusters  with
                         proper  intensity ratios must  also be present.
    
                11.2.3    Alternatively, a spectral search may be used to  auto-
                         matically  reduce the data.  The criteria  for acceptable
                         identification include a  high index of similarity.   For
                         the -Incos  2300, a  fit of  750  or greater must be  obtained.
    
          11.3  Disputes in interpretation  - Where there is  reasonable doubt  as
                to the  identity  of  a peak as a PCB,  the  analyst must  either iden-
                tify the peak as a  PCB  or proceed  to a confirmational analysis
                (see Section 13.0).
    
    
    12.0  Quantitative  Data Reduction
    
          12.1  Once a  chromatographic  peak has been identified as a  PCB, the com-
                pound is quantitated based  either  on the integrated abundance of
                the SIM data or  EICP for the primary characteristic ion in  Tables
                11 and  12.  If interferences are observed for the  primary ion,
                use the secondary and then  tertiary  ion  for  quantitation.   If
                interferences in the parent cluster  prevent  quantitation, an  ion
                from a  fragment  cluster (e.g., M-70) may be  used.   Whichever  ion
                is used, the RF  must be determined using that ion. The same  cri-
                teria should be  applied to  the surrogate compounds (Table 13).
    
          12.2  Using the appropriate analyte-internal standard pair  and  response
                factor  (RF ) as  determined  in Section  7.3, calculate  the  concen-
                tration ofpeach  peak using  Equation  12-1.
    
                                             A)     i     M      V
                      Concentration (fjg/g)  = -E_ . ^- . gi£ .  ^£       Eq. 12-1
                                              is    p   e     i
    
                where     A  = area of  the  characteristic ion for  the analyte PCB
                           P    peak
    
                         A.  = area of  the  characteristic ion for  the internal
                                standard peak
    
                         RF  = response factor of  a  given PCB congener
    
    
                                       B-33
    

    -------
            TABLE 13.  CHARACTERISTIC IONS FOR 13C-LABELED PCB SURROGATES
    
    
    
                                   "	Ion (relative intensity)	
    Specific compound              Primary           Secondary          Tertiary
    
    
    
    
    
    
    13C612C6H9C1                  194 (100)          196 (33)
    
    
    
    
    13C12H6C14                    304 (100)          306 (49)           302  (78)
    
    
    
    
    13C12H2C18                    442 (100)          444 (65)           440  (89)
    
    
    
    
    13C12C110                     510 (100)          512 (87)           514  (50)
                                         B-34
    

    -------
                   M.   = mass of internal standard injected (micrograms)
                    XS   ( , •
    
                    M  = mass of sample extracted (grams)
    
                    V. = volume injected (microliters)
    
                    V  = volume of sample extract (microliters)
    
    12.3  If a peak appears  to contain non-PCB interferences,  which cannot
          be circumvented by a secondary or tertiary ion,  either:
    
          12.3.1   Reanalyze the sample on a different column which sepa-
                   rates the PCB and interf erents ;
    
          12.3.2   Perform additional chemical cleanup (Section 9) and then
                   reanalyze the sample; or
    
          12.3.3   Quantitate the entire peak as PCB.
    
    12.4  Calculate the recovery of the four 13C surrogates using  the ap-
          propriate surrogate-internal standard pair and response  factor
          (RF. ) as determined in Section 7.4 using Equation 12-2.
             xs
                                A           M.
                 Recovery (%)=-£_. ^- . -*i . 100             Eq. 12-2
                                 IS     S    S
          where A  = area of the characteristic ion for the surrogate peak
                 S
    
               A.  = area of the characteristic ion for the internal standard
                is        ,
                       peak
    
               RF  = response factor for the surrogate compound with respect
                       to the internal standard (Equation 7-2)
    
               M.  = mass of internal standard injected (nanograms)
                XS
                M  = mass of surrogate, assuming 100% recovery (nanograms)
                 S
    
    12.5  Correct the concentration of each peak using Equation 12-3.  This
          is the final reportable concentration.
      Corrected concentration (pg/g) =                    *  10°    ^  12'3
    12.6  Sum all of the peaks for each homolog,  and then sum those to yield
          the total PCB concentration in the sample.  Report all numbers in
          Mg/g.  The reporting form in Table 14 may be used.  If an alter-
          nate reporting format (e.g., concentration per peak) is desired,
          a different report form may be used.   The uncorrected concentra-
          tions, percent recovery, and corrected recovery are to be reported.
    
    12.7  Round off all numbers reported to two significant figures.
    
    
                                 B-35
    

    -------
                              TABLE 14.  ANALYSIS REPORT
              INCIDENTAL PCBs IN COMMERCIAL PRODUCTS OR PRODUCT WASTES
    Sample No. 	
    Sample Matrix	
    Sample Source 	
    Notebook No. or File Location
    
    Volume Extracted 	
    Extraction/Cleanup Procedure
    Int. Std.
    4-Cl(d6)
       Mass Added (|Jg)
         (Circle one)
    
          298    300
            Ratio
                                             100/49
             Intensity
    Surrogates
    
      1-C1
    
      4-C1
    
      8-C1
    
     10-C1
    Mass Added (|Jg)
    (Circle one)
    
     194    196
    
     304    306
    
     442    444
    
     510    512
     Ratio
    
    100/33
    
    100/49
    
    100/65
    
    100/87
    Intensity   % Recovery
                                     (continued)
                                        B-36
    

    -------
    TABLE 14 (continued)
    
    
    Analyte 1° 2° Il°
    1-C1 188 190
    2-C1 222 224
    3-C1 256 258
    4-C1 292 290
    5-C1 326 328
    6-C1 360 362
    7-C1 394 396
    8-C1 430 432
    9-C1 464 466
    10-C1 498 500
    Total
    Reported by:
    Name
    Signature/Date
    Organization
    Qualitative
    2° Ratio Theoretical
    100/33
    100/66
    100/99
    100/76
    100/66
    100/82
    100/98
    100/66
    100/76
    100/87
    Internal Audit:
    Name
    Signature/Date
    Organization
    Quantitative
    Uncorr. Corr.
    Ion Cone. Cone.
    OK? Used RF (jjg/g) (Mg/g)
    
    
    
    
    
    
    
    
    
    
    Mg/g Mg/g
    Uncorr. Corr.
    EPA Audit:
    Name
    Signature/Date
    Organization
       B-37
    

    -------
    13.0  Confirmation
    
          If there is reason to question the qualitative identification (Section
          11.0),  the analyst may choose to confirm that a peak is not a PCB.   Any
          technique may be chosen provided that it is validated as having equiva-
          lent or superior selectivity and sensitivity to GC/EIMS.  Some candidate
          techniques include alternate GC columns  (with EIMS detection), GC/CIMS,
          GC/NCIMS, high resolution EIMS, and MS/MS techniques.   Each laboratory
          must validate confirmation techniques to show equivalent or superior
          selectivity between PCBs and interferences and sensitivity (limit of
          quantitation, LOQ).
    
          If a peak is confirmed as being a non-PCB, it may be deleted from the
          calculation (Section 12).  If a peak is  confirmed as containing both
          PCB and non-PCB components, it must be quantitated according to Section
          12.3.
    14.0  Quality Control
    
          14.1  Each laboratory that uses this method must operate a formal qual-
                ity control (QC) program.  The minimum requirements of this pro-
                gram consist of an initial demonstration of laboratory capability
                and the analysis of spiked samples as a continuing check on perfor-
                mance.   The laboratory must maintain performance records to define
                the quality of data that are generated.  After a date specified by
                the Agency, ongoing performance checks should be compared with
                established performance criteria to determine if the results of
                analyses are within accuracy and precision limits expected of the
                method.
    
          14.2  The analysts must certify that the precision and accuracy of the
                analytical results are acceptable by:
                     (
                14.2.1    The absolute precision of surrogate recovery, measured
                         as the RSD  of the integrated EIMS area (A ) for a set
                         of samples, must be ±10%.
    
                14.2.2    The mean recovery (R ) of at least four replicates of a
                         QC check sample to be supplied by the Agency must meet
                         Agency-specified accuracy and precision criteria.  This
                         forms the initial data base for establishing control
                         limits (see Section 14.3 below).
    
          14.3  Control limits - The laboratory must establish control limits
                using the following equations:
    
                Upper control limit (UCL) = R  +3 RSD
    
                Upper warning limit (UWL) = R  +2 RSD
                Lower warning limit (LWL)  = R  - 2 RSD
    
                Lower control limit (LCL)  = RC - 3 RSDc
    
                                       B-38
    

    -------
          These may be plotted on control charts.   If an analysis  of a check
          sample falls outside the warning limits,  the analyst should be
          alerted that potential problems may need  correction.  If the re-
          sults for a check sample fall outside the control limits, the lab-
          oratory must take corrective action and recertify the performance
          (Section 14.2)  before proceeding with analyses.   The warning and
          control limits  should be continuously updated as more check sample
          replicates are  added to the data base.
    
    14.4  Before processing any samples, the analyst should demonstrate
          through the analysis of a reagent blank that all glassware and
          reagent interferences are under control.   Each time a set of sam-
          ples is analyzed or there is a change in  reagents, a laboratory
          reagent blank should be processed as a safeguard against contam-
          ination.
    
    14.5  Procedural QC - The various steps of the  analytical procedure
          should have quality control measures. These include but are not
          limited to:
    
          14.5.1   GC performance - See Section 7.1 for performance criteria.
    
          14.5.2   MS performance - See Section 7.2 for performance criteria.
    
          14.5.3   Qualitative identification - At  least 10% of the PCB
                   identifications, as well as any  questionable results,
                   should be confirmed by a second  mass spectrometrist.
    
          14.5.4   Quantitation - At least 10% of all manual calculations,
                   including peak area calculations, must be checked.  After
                   changes in computer quantitation routines, the  results
                   should be manually checked.
    
    14.6  A minimum of 10% of all samples, one sample per month or one sam-
          ple per matrix type, whichever is greater, selected at random, must
          be run in triplicate to monitor the precision of the analysis.  An
          RSD of ±30% or less must be achieved.  If the precision  is greater
          than ±30%, the analyst must be recertified (see Section  14.2).
    
    14.7  A minimum of 10% of all samples, one sample per month or one sam-
          ple per matrix type, whichever is greater, selected at random, must
          be analyzed by the standard addition technique.   Two aliquots of
          the sample are analyzed, one "as is" and  one spiked (surrogate
          spiking and equilibration techniques are  described in Section 9.2)
          with a sufficient amount of Solution CSxxx to yield approximately
          100 pg/g of each compound.  The samples are analyzed together and
          the quantitative results calculated.  The recovery of the spiked
          compounds (calculated by difference) must be 80-120%.  If the sam-
          ple is known to contain specific PCB isomers, these isomers may be
          substituted for solution CSxxx.  If the concentrations of PCBs are
          known to be high or low, the amount added should be adjusted so
          that the spiking level is 1.5 to 4 times  the measured PCB level
          in the unspiked sample.
    
                                 B-39
    

    -------
          14.8  Interlaboratory comparison -  Interlaboratory comparison  studies
                are planned.   Participation requirements,  level  of  performance,
                and the identity of the coordinating laboratory  will  be  presented
                in later revisions.
    
          14.9  It is recommended that the participating  laboratory adopt  addi-
                tional QC practices for use with this method.  The  specific  prac-
                tices that are most productive  depend upon the needs  of  the  lab-
                oratory and the nature of the samples.  Field duplicates or  trip-
                licates may be analyzed to monitor the precision of the  sampling
                technique.  Whenever possible,  the laboratory should  perform
                analysis of standard reference  materials  and participate in  rele-
                vant performance evaluation studies.
    
    
    15.0  Quality Assurance
    
          Each participating laboratory must develop a quality assurance plan ac-
          cording to EPA guidelines.5  The quality assurance plan must be  submitted
          to the Agency for approval.
    
    
    16.0  Method Performance
    
          The method performance is being evaluated.   Limits of  quantitation;
          average intralaboratory recoveries, precision,  and accuracy; and inter-
          laboratory recoveries, precision, and accuracy  will be presented.
    
    
    17.0  Documentation and Records
    
          Each laboratory is responsible for maintaining  full records of the analy-
          sis.  Laboratory notebooks should be  used for handwritten records.   GC/MS
          data must be archived on magnetic tape, disk, or a similar  device.   Hard
          copy printouts may be kept in addition if desired. QC records should
          be maintained separately from sample  analysis records.
    
          The documentation must describe completely how  the analysis was  performed.
          Any variances from the protocol must  be noted and fully described.   Where
          the protocol lists options (e.g., sample cleanup), the option  used and
          specifics (solvent volumes, digestion times, etc.) must be  stated.
                                       B-40
    

    -------
                                     REFERENCES
    
    1.   "Methods 330.4 (Titrimetric,  DPD-FAS)  and 330.5  (Spectrophotometric,  DPD)
        for Chlorine,  Total Residual," Methods for Chemical Analysis  of Water and
        Wastes, U.S.  Environmental Protection  Agency,  Environmental Monitoring
        and Support Laboratory,  Cincinnati,  Ohio, March  1979,  EPA 600-4/79-020.
    
    2.   Erickson, M.  D.,  and J.  S. Stanley,  "Methods of  Analysis for  Incidentally
        Generated PCBs—Literature Review and  Preliminary Recommendations," Interim
        Report No. 1,  EPA Contract No. 68-01-5915, Task  51, 1982.
    
    3.   Bellar, T. A., and J. J.  Lichtenberg,  "The Determination of Polychlorinated
        Biphenyls in Transformer Fluid and Waste Oils,"  Prepared for  U.S.  Environ-
        mental Protection Agency, (1981) EPA-600/4-81-045.
    
    4.   American Society for Testing and Materials, "Standard  Method  for Analysis
        of Environmental Materials for Polychlorinated Biphenyls," pp.  877-885 in
        Annual Book of ASTM Standards, Part  40, Philadelphia,  Pennsylvania (1980).
        ANSI/ASTM D 3304 - 77.
    
    5.   "Quality Assurance Program Plan for  the Office of Toxic Substances,"
        Office of Pesticides and Toxic Substances, U.S.  Environmental Protection
        Agency, Washington, B.C., October 1980.
                                       B-41
    

    -------
                      APPENDIX C
    
    
    ANALYTICAL METHOD:  THE ANALYSIS OF BY-PRODUCTS
             CHLORINATED BIPHENYLS IN AIR
                        C-l
    

    -------
               THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS IN AIR
    1.0   Scope and Application
    
          1.1   This is a gas chromatographic/electron impact mass spectrometric
                (GC/EIMS) method applicable to the determination of chlorinated
                biphenyls (PCBs) in air emitted from commercial production through
                stacks, as fugitive emissions, or static (room, other containers,
                or outside)  air.  The PCBs present may originate either as syn-
                thetic by-products or as contaminants derived from commercial PCB
                products (e.g., Aroclors).  The PCBs may be present as single
                isomers or complex mixtures and may include all 209 congeners
                from monochlorobiphenyl through decachlorobiphenyl listed in
                Table 1.
    
          1.2   The detection and quantitation limits are dependent upon the vol-
                ume of sample collected, the complexity of the sample matrix and
                the ability of the analyst to remove interferents and properly
                maintain the analytical system.  The method accuracy and preci-
                sion will be determined in future studies.
    
          1.3   This method is restricted to use by or under the supervision of
                analysts experienced in the use of gas chromatography/mass spec-
                trometry (GC/MS) and in the interpretation of gas chromatograms
                and mass spectra.  Prior to sample analysis, each analyst must
                demonstrate the ability to generate acceptable results with this
                method by following the procedures described in Section 14.2.
    
          1.4   The validity of the results depends on equivalent recovery of the
                analyte and 13C PCBs.  If the 13C PCBs are not thoroughly incor-
                porated in the matrix, the method is not applicable.
    
          1.5   During the development and testing of this method, certain an-
                alytical parameters and equipment designs were found to affect
                the validity of the analytical results.  Proper use of the method
                requires that such parameters or designs must be used as speci-
                fied.  These items are identified in the text by the word "must."
                Anyone wishing to deviate from the method in areas so identified
                must demonstrate that the deviation does not affect the validity
                of the data.  Alternative test procedure approval must be ob-
                tained from the Agency.  An experienced analyst may make modifi-
                cations to parameters or equipment identified by the term "recom-
                mended."  Each time such modifications are made to the method,
                the analyst must repeat the procedure in Section 14.2.  In this
                case, formal approval is not required, but the documented data
                from Sectin 14.2 must be on file as part of the overall quality
                assurance program.
                                       C-2
    

    -------
                                   TABLE 1.  NUMBERING OF PCB  CONGENERS3
    NO.
    
    1
    3
    
    
    
    4
    5
    6
    7
    a
    9
    10
    11
    12
    13
    14
    15
    
    
    
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    26
    27
    23
    29
    30
    31
    32
    33
    34
    35
    36
    37
    38
    39
    
    
    
    40
    41
    42
    43
    44
    45
    46
    47
    48
    49
    50
    51
    
    Structure
    ManochloroblphtriYU
    2
    4
    
    D1cMoroMpti«ny1>
    
    2.2'
    2,3
    2.3'
    2.4
    2.4'
    .5
    ,6
    ,3'
    ,4
    ,4'
    .5
    •».«'
    
    TrtchlorolilphenyU
    
    2.2', 3
    2.2'. 4
    2.2'. 5
    2.2'. 6
    2,3,3'
    2.3,4
    2.3.4'
    2.3.5
    2.3,6
    2,3', 4
    2.3'. 5
    2, 3', 6
    2,4,4'
    2.4.5
    2,4.6
    2. 4'. 5
    2,4', 6
    2'. 3. 4
    2'. 3,5
    3.3',4
    3.3',5
    3.4,4'
    3.4.5
    3. 4'. 5
    
    Tttracft 1 oroftl oh«nyl s
    
    2,2' ,3,3'
    2,2'. 3.4
    2.2', 3,4'
    2.2'.3,5
    2,2',3,5'
    2,2', 3, 6
    2.2'. 3,6'
    2.2', 4,4'
    2.2', 4,5
    2.2', 4.5'
    2.2', 4,6
    2,2', 4.6'
    
    NO.
    
    52
    S4
    55
    56
    57
    58
    59
    60
    61
    62
    63
    M
    65
    66
    67
    68
    69
    70
    71
    72
    73
    7«
    75
    76
    77
    78
    79
    80
    31
    
    
    
    82
    83
    84
    85
    86
    87
    88
    99
    90
    91
    92
    93
    9*
    95
    96
    97
    98
    99
    100
    101
    102
    103
    104
    
    
    
    
    
    structurt
    T«tricftlofnb1ph»nylt
    2.2'. 5,5'
    29* e ic*
    • & |9*O
    2.2'. 6,6'
    2,3. 3'N
    2.3,3',4'
    2.3,3' .S
    2.3.3'. 5'
    2.3,3', 6
    2.3,4,4'
    2,3,4.5
    2,3.4,6
    2.3,4'.S
    2. 3,4', 6
    2.3,5.6
    2.3', 4,4'
    2, 3', 4,5
    2.3', 4,5'
    2.3'. 4,6
    2,3'.V.S
    2,3',4>,6
    2.3' ,5.5'
    2. 3', 5', 6
    2.4, 4'. S
    2.4.41.6
    2'. 3 4,$
    3.3' 4.4'
    3.3' 4,5
    3.3' 4.5'
    3.3' 5,5'
    3.4. 41, 5
    
    P«ntidi1oro61phenyl«
    
    2,2',3,3',4
    2. 2', 3, 3', 5
    2.2' .3, 3' ,6
    2.2' .3,4,4'
    2.2',3,4.5
    2,2'.3,4.5'
    2.2', 3,4,6
    2.2'. 3,4, 6'
    2.2'.3.4'.5
    2, 2'. 3,4'. 6
    2.2' ,3.5. 5'
    2,2'. 3, 5, 6 •>
    2.2'. 3,5,6'
    2,2'. 3.5', 8
    2.2',3.6.6'
    2.2' ,3' .4,5
    2,2' ,3'. 4,6
    2,2',4,4'15
    2,2'. 4. 4' .6
    2.2', 4,5.5'
    2,2'.4.5,6'
    2,2'. 4,5'. 6
    2,2' .4.6.6'
    
    
    
    
    
    No.
    
    105
    1/VE
    IUD
    107
    108
    109
    110
    111
    112
    113
    114
    115
    116
    117
    118
    119
    120
    121
    122
    123
    124
    125
    126
    127
    
    
    
    128
    129
    130
    131
    132
    133
    134
    135
    136
    137
    138
    139
    140
    141
    142
    143
    144
    145
    146
    147
    148
    149
    150
    151
    152
    153
    154
    155
    156
    157
    158
    159
    160
    
    
    Structure
    Pentich 1 orob t ohefiy 1 s
    2.3,3'. 4,4'
    24 ^, 1 t
    ,3,3 .4,9
    2, 3,3' .4'. 5
    2,3,3', 4. 5'
    2.3, 3', 4, 6
    2. 3, 3' ,4', 6
    2.3,3' ,5,5'
    2, 3, 3', 5,6
    2.3.3'. 5'.6
    2.3,4. 41 .5
    2.3.4,4'.$
    2.3.4.5.6
    2.3.4' ,5,6
    2.3'.4.4'.S
    2.3'. 4,4'. 6
    2,3', 4,5, 5'
    2,3', 4. 5', 6
    21. 3,3'. 4.5
    2'. 3. 4. 4'. 5
    2'. 3. 4,5. 5'
    2'.3.4.5.6'
    3, 3', 4. 4'. 5
    3, 3', 4, 5, 5'
    
    Htmehlorabfphenyli
    
    2, 2', 3, 3'. 4,4'
    2.2', 3, 3' .4.5
    2.2', 3. 3'. 4, 5'
    2.2'.3.3',4.6
    2.2', 3.3'. 4,6'
    2. 2'. 3, 3'. 5, 5'
    2, 2', 3, 3', 5, 6
    2.2', 3, 3'. 5, 6'
    2, 2', 3, 3'. 6, 6'
    2.2' .3,4,4' ,5
    2,2'. 3, 4, 4'. 5'
    2.2' .3,4, 4' ,6
    2,2', 3, 4,4'. 6'
    2.2',3.4,5,5'
    2.2'. 3, 4, 5,6
    2.2',3.4.5.6'
    2.2'.3,4.5',6
    2.2'.3,4,6.6'
    2.2'. 3.4'. 5.5'
    2.2'.3,4',5.6
    2.2'. 3. 4'. 5. 6'
    2.2'. 3,4'. 5' .6
    2,2'. 3,4'. 6.6'
    2.2'.3,S,5',6
    2.2'. 3,5.6,6'
    2.2' ,4, 4', 5.5'
    2,2'. 4,4', 5.6'
    2.2'. 4, 4'. 6,6'
    2.3,3'. 4,4' ,5
    2.3. 3' .4, 4'. 5'
    2,3,3'.4.4',6
    2.3,3' .4.5. 5'
    2.3,3'. 4,5,6
    
    
    NO.
    
    161
    1X9
    IO*
    163
    164
    165
    166
    167
    168
    169
    
    
    
    170
    171
    172
    173
    174
    175
    176
    177
    178
    179
    180
    181
    182
    183
    184
    185
    186
    187
    188
    189
    190
    191
    192
    193
    
    
    
    194
    195
    196
    197
    198
    199
    200
    201
    202
    203
    204
    205
    
    
    
    206
    207
    208
    
    
    
    209
    structure
    Hemch! orobl phenyl i
    2. 3. 3'. 4. 5'. 6
    2.3,3' ,4'. 5,6
    2, 3, 3'. 4'. 5' .6
    2.3.3', 5, 5'. 6
    2, 3, 4, 4'. 5, 6
    2.3'. 4, 4'. 5. 5'
    2. 3' ,4, 4'. 5'. 6
    3,3' .4. 4'. 5,5'
    
    Hetmchlorobtphenyl t
    
    2.2' .3. 3'. 4. 4', 5
    2, 2'. 3, 3' ,4. 41. 6
    2. 2'. 3. 3'. 4. 5, 5'
    2, 2', 3, 3', 4, 5, 6
    2, 2'. 3. 3'. 4, 5, 6'
    2.2'.3,3'.4,51,6
    2,2'.3.3'.4.6.6'
    2, 2'. 3, 3'. 4'. 5, 6
    2, 2', 3. 3', 5, 5'. 6
    
    
    
    2, 2', 3.4, 4'. 5. 6'
    2. 2'. 3,4,4'. 5'. 5
    2. 2'. 3, 4, 4'. 6, 6'
    2,2' ,3, 4.5, 5'. 6
    2, 2'. 3, 4, 5. 6, 6'
    2, 2', 3, 4', 5. 5'. 6
    2. 2', 3. 41. 5, 6, 6'
    2. 3, 3', 4. 4'. 5, 5'
    2, 3, 3'. 4, 4', 5. 6
    2, 3, 3', 4, 4' ,5', 5
    2, 3, 3', 4, 5,5', 6
    2.3,3',4'.5,51,S
    
    OctiCBlorobipftenyl i
    
    2,2' .3. 3' .4, 4', 5, 5'
    2,2', 3, 3', 4, 4'. 5,0
    2,2' .3, 3', 4, 4'. 3,6'
    2. 2'. 3, 3', 4, 4', 6, 6'
    2.2'. 3, 3' ,4, 5, 5' ,o
    2. 2', 3. 3'. 4, 5, 6, 6'
    2,2',3,3'.4.5',6.6'
    2.2'. 3, 3', 4. 5, 5'. 6'
    2.2* .3.3' .5,5', 6,6'
    2,2' .3,4,4'. 5,5',6
    2.2'. 3,4, 4'. 5, 6, 6'
    2.3.3'.4,4'.5,5',6
    
    Honieh 1 orob< oneny 1 ?
    
    2.2'.3,3'.4,4',S.5'.6
    2, 2' ,3, 3' .4, 4', 5, 6,6'
    2,2'. 3. 3'. 4, 5, 5'. 6. 6'
    
    Oecacft 1 orom oheny 1
    
    Z^'.J.JU.i'.s.s'.e.s1
    •Adopttd fro» BtllscMwr. X. ind 2*11, M., Fr««n1ui Z.  Anal.Omi., 302. 20-31  (1980).
                                                 C-3
    

    -------
    2.0   Summary
          2.1   The air must be sampled such that the specimen collected for
                analysis is representative of the whole.   Statistically designed
                selection of the sampling position (stack, flue, port, etc.) or
                time should be employed.  Gaseous and particulate PCBs are with-
                drawn isokinetically from stacks, room air exhausts, process point
                exhausts, and other flowing gaseous streams using a sampling train.1
                The PCBs are collected in the Florisil adsorbent tube and in the
                impingers in front of the adsorbent.   PCBs are sampled from ambient
                air and other static gaseous sources  onto a Florisil adsorbent
                tube.  The sample must be preserved to prevent PCB loss prior to
                analysis.  Storage at 4°C is recommended.
    
          2.2   The Florisil adsorbent is extracted with hexane in a Soxhlet ex-
                tractor, the aqueous condensate is extracted with hexane and the
                acetone/hexane impinger rinse is back-extracted with water.  All
                three organic extracts are then combined.  Optional cleanup tech-
                niques may include sulfuric acid cleanup and Florisil adsorption
                chromatography.  The sample is concentrated to a final known vol-
                ume for instrumental determination.
    
          2.3   The PCB content of the sample extract is determined by capillary
                (preferred) or packed column gas chromatography/electron impact
                mass spectrometry (CGC/EIMS or PGC/EIMS) operated in the selected
                ion monitoring (SIM), full scan, or limited mass scan (LMS) mode.
    
          2.4   PCBs are identified by comparison of their retention time and mass
                spectral intensity ratios to those in calibration standards.
    
          2.5   PCBs are quantitated against the response factors for a mixture
                of 11 PCB congeners using the internal standard technique.
    
          2.6   The PCBs identified by the SIM technique may be confirmed by full
                scan CGC/EIMS, retention on alternate GC columns, other mass spec-
                trometric techniques, infrared spectrometry, or other techniques,
                provided that the sensitivity and selectivity of the technique are
                demonstrated to be comparable or superior to GC/EIMS.
    
          2.7   The analysis time is dependent on the extent of workup employed.
                The time required for instrumental analysis of a single sample
                excluding data reduction and reporting, is about 30 to 45 min.
    
          2.8   Appropriate quality control (QC) procedures are included to assess
                the performance of the analyst and estimate the quality of the re-
                sults .  These QC procedures include the demonstration of laboratory
                capability:  periodic analyst certification, the use of control
                charts, and the analysis of blanks, replicates, and standard addi-
                tion samples.  A quality assurance (QA) plan must be developed for
                each laboratory.
                                       C-4
    

    -------
          2.9   While several options are available throughout this  method,  the
                recommended procedure for stack gases to be followed is:
    
                2.9.1    The sample is collected using a modified Method  5 train1
                         according to a scheme which permits extrapolation of the
                         sample data to the source being assessed.
    
                2.9.2    The sample is preserved at 4°C to prevent any loss  of
                         PCBs or changes in matrix which may adversely affect re-
                         covery.
    
                2.9.3    The three sample fractions are extracted and combined.
    
                2.9.4    The extract is cleaned up and concentrated  to an appro-
                         priate volume.  Internal standards are added.
    
                2^9.5    An aliquot of the extract is analyzed by CGC/EIMS oper-
                         ated in the SIM mode.  On-column injections onto a  15-m
                         DB-5 capillary column, programmed (for toluene solutions)
                         from 110° to 325°C at 10°/min after a 2 min hold is used.
                         Helium at 45-cm/sec linear velocity is used as the  car-
                         rier gas.
    
                2.9.6    PCBs are identified by retention time and mass spectral
                         intensities.
    
                2.9.7    PCBs are quantitated against the response factors for a
                         mixture of 11 PCB congeners.
    
                2.9.8    The total PCBs are obtained by summing the amounts  for
                         each homolog found, and the concentration is reported
                         as micrograms per cubic meter.
    
    
    3.0   Interferences
    
          3.1   Method interferences may be caused by contaminants,  in sample col-
                lection media, solvents, reagents, glassware, and other sample
                processing hardware, leading to discrete artifacts and/or ele-
                vated baselines in the total ion current profiles.   All of these
                materials must be routinely demonstrated to be free  from interfer-
                ences by the analysis of laboratory reagent blanks as described
                in Section 14.4.
    
                3.1.1    Glassware must be scrupulously cleaned.  All glassware
                         is cleaned as soon as possible after use by rinsing with
                         the last solvent used.  This should be followed by deter-
                         gent washing with hot water and rinses with tap water and
                         reagent water.  The glassware should then be drained dry
                         and heated in a muffle furnace at 400°C for 15 to 30 min.
                         Some thermally stable materials, such as PCBs, may not
                         be eliminated by this treatment.  Solvent rinses with
                         acetone and pesticide quality hexane may be substituted
    
                                       C-5
    

    -------
                         for the muffle furnace heating.   Volumetric ware should
                         not be heated in a muffle furnace.   After it is dry and
                         cool,  glassware should be sealed and stored in a clean
                         environment to prevent any accumulation of dust or other
                         contaminants.  It is stored inverted or capped with
                         aluminum foil.
    
                3.1.2    The use of high purity reagents  and solvents helps to
                         minimize interference problems.   Purification of solvents
                         by distillation in all-glass systems may be required.
                         All solvent lots must be checked for purity prior to use.
    
          3.2   Matrix interferences may be caused by contaminants that are coex-
                tracted from the sorbent material or impingers.   The extent of
                matrix interferences will vary considerably  from source to source,
                depending upon the nature and diversity of the sources of samples.
    4.0   Safety
          4.1   The toxicity or carcinogenicity of each reagent used in this
                method has not been precisely defined;  however, each chemical
                compound should be treated as a potential health hazard.   From
                this viewpoint, exposure to these chemicals must be reduced to
                the lowest possible level by whatever means available.   The lab-
                oratory is responsible for maintaining a current awareness file
                of OSHA regulations regarding the safe handling of the  chemical
                specified in this method.  A reference file of material data han-
                dling sheets should also be made available to all personnel in-
                volved in the chemical analysis.
    
          4.2   Polychlorinated biphenyls have been tentatively classified as
                known or suspected human or mammalian carcinogens.  Primary
                standards of these toxic compounds should be prepared in a hood.
                Personnel must wear protective equipment, including gloves and
                safety glasses.
    
                Congeners highly substituted at the meta and para positions and
                unsubstituted at the ortho positions are reported to be the most
                toxic.  Extrme caution should be taken when handling these com-
                pounds neat or in concentrated solution.  The class includes
                3,3',4'4'-tetrachlorobiphenyl (both natural abundance and isotop-
                ically labeled).
    
          4.3   Waste disposal must be in accordance with RCRA and applicable
                state rules.
    5.0   Apparatus and Materials
    
          All specifications are suggestions only.   Catalog numbers and suppliers
          are included for illustration only.
    
    
                                       C-6
    

    -------
    5.1   Stack sampling train1 - See Figure 1; a series of four impingers
          with a solid adsorbent trap between the third and fourth impingers.
          The train may be constructed by adaptation from a Method 5 train.2
          Descriptions of the train components are contained in the follow-
          ing subsections.
    
          5.1.1    Probe nozzle - Stainless steel (316) with sharp, tapered
                   leading edge.  The angle of taper shall be ^ 30° and the
                   taper shall be on the outside to preserve a constant in-
                   ternal diameter.  The probe nozzle shall be of the button-
                   hook or elbow design, unless otherwise specified by the
                   Agency.  The wall thickness of the nozzle shall be less
                   than or equal to that of 20 gauge tubing, i.e., 0.165 cm
                   (0.065 in.) and the distance from the tip of the nozzle
                   to the first bend or point of disturbance shall be at
                   least two times the outside nozzle tubing.  Other con-
                   figurations and construction material .may be used with
                   approval from the Agency.
    
          5.1.2    Probe liner - Borosilicate or quartz glass equipped with
                   a connecting fitting that is capable of forming  a leak-
                   free, vacuum tight connection without sealing greases;
                   such as Kontes Glass Company "0" ring spherical ground
                   ball joints (model K-671300) or University Research
                   Glassware SVL teflon screw fittings.
    
                   A stainless steel (316) or water-cooled probe may be used
                   for sampling high temperature gases with approval from
                   the Agency.  A probe heating system may be used to prevent
                   moisture condensation in the probe.
    
          5.1.3    Pitot tube - Type S, or equivalent, attached to probe to
                   allow constant monitoring of the stack gas velocity.
                   The face openings of the pitot tube and the probe nozzle
                   shall be adjacent and parallel to each other but not
                   necessarily on the same plane, during sampling.  The free
                   space between the nozzle and pitot tube shall be at least
                   1.9 cm (0.75 in.).  The free space shall be set based on
                   a 1.3 cm (0.5 in.) ID nozzle, which is the largest size
                   nozzle used.
    
                   The pitot tube must also meet the criteria specified in
                   Method 22 and be calibrated according to the procedure
                   in the calibration section of that method.
    
          5.1.4    Differential pressure gauge - Inclined manometer capable
                   of measuring velocity head to within 10% of the minimum
                   measured value.  Below a differential pressure of 1.3 mm
                   (0.05 in.) water gauge, micromanometers with sensitivities
                   of 0.013 mm (0.0005 in.) should be used.  However, micro-
                   manometers are not easily adaptable to field conditions
                   and are not easy to use with pulsating flow.  Thus, other
                   methods or devices acceptable to the Agency may be used
                   when conditions warrant.
    
                                 C-7
    

    -------
                            Stack
                            Wall
                                                         Thermometer'
    
                                                Florisil Tube
    Check
    Valve
            Probe (r^
                  £
    Reverse-Type
    Pitor Tube
                                             Control Box
                          Figure 1.  PCB sampling train for  stack  gases.
                                               C-J
    

    -------
          5.1.5    Impingers - Four impingers with connecting fittings able
                   to form leak-free,  vacuum tight seals without sealant
                   greases when connected together as shown in Figure 1.
                   The first and second impingers are of the Greenburg-
                   Smith design.  The  final two impingers are of the
                   Greenburg-Smith design modified by replacing the tip
                   with a 1.3 cm (1/2  in.) ID glass tube extending to 1.3
                   cm (1/2 in.) from the bottom of the flask.
    
                   One or two additional modified Greenburg-Smith impingers
                   may be added to the train between the third impinger and
                   the Florisil tube to accommodate additional water col-
                   lection when sampling high moisture gases.  Throughout
                   the preparation, operation, and sample recovery from the
                   train, these additional impingers should be treated
                   exactly like the third impinger.
    
          5.1.6    Solid adsorbent tube - Glass with connecting fittings
                   able to form leak-free, vacuum tight seals without seal-
                   ant greases (Figure 2).  Exclusive of connectors, the
                   tube has a 2.2 cm inner diameter, is at least 10 cm long,
                   and has four deep indentations on the inlet end to aid
                   in retaining the adsorbent.  Ground glass caps (or
                   equivalent) must be provided to seal the adsorbent-filled
                   tube both prior to  and following sampling.
    
          5.1.7    Metering system - Vacuum gauge, leak-free pump, thermom-
                   eters capable of measuring temperature to within ±3°C
                   (•^ 5°F), dry gas meter with 2% accuracy at the required
                   sampling rate, and  related equipment, or equivalent, as
                   required to maintain an isokinetic sampling rate and to
                   determine sample volume.  When the metering system is
                   used in conjunction with a pitot tube, the system shall
                   enable checks of isokinetic rates.
                           i
    
          5.1.8    Barometer - Mercury, aneroid, or other barometers cap-
                   able of measuring atmospheric pressure to within 2.5 mm
                   Hg (0.1 in. Hg). In many cases, the barometric reading
                   may be obtained from a nearby weather bureau station, in
                   which case the station value shall be requested and an
                   adjustment for elevation differences shall be applied at
                   a rate of -2.5 mm Hg (0.1 in. Hg) per 30 mm (100 ft) ele-
                   vation increase.
    
    5.2   Static air sampling train1 - The sampling train, see Figure 3,
          consists of a glass-lined probe, an adsorbent tube containing
          Florisil, and the appropriate valving and flow meter controls for
          isokinetic sampling as described in Section 5.1.  The sampling
          apparatus in Figure 3 is the same as that in Figure 1 and Section
          5.1, except that the Smith-Greenburg impingers and heated probe
          are not used.  If condensation of significant quantities of mois-
          ture prior to the solid adsorbent is expected, Section 5.1 of the
                                 C-9
    

    -------
                               J28/12
    10cm
                              j 28/12
    Figure 2.   Florisil adsorbent tube.
                   C-10
    

    -------
                           Probe (to sample from duct) ^—
                                   Glass- lined Probe
        a
    Type S
    Pitot Tube
                                          Manometer
                        Thermometers
                         0(7
                        By - pass
                        Valve
                                                                    Florisil
    
                                                                    Glass Wool
                                                                     Check Valve
    Vacuum
    Gauge
                       I  Integrated  |
                         Flow Meter
    Manometer  -
                               Air
                               Tight
                               Pump
    Vacuum
    Line
                   Figure  3.   PCB  sampling  train  for  static air.
                                       C-ll
    

    -------
          method should be used.   Since probes and adsorbent tubes are not
          cleaned up in the field, a sufficient number must be provided for
          sampling and allowance for breakage.
    
    5.3   Sample recovery
    
          5.3.1    Ground glass caps - To cap off adsorbent tube and the
                   other sample exposed portions of the train.
    
          5.3.2    Teflon FEP® wash bottle - Two, 500 ml, Nalgene No.
                   0023A59 or equivalent.
    
          5.3.3    Sample storage containers - Glass bottles, 1 liter, with
                   TFE®-lined screw caps.
    
          5.3-4    Balance - Triple beam, Ohaus Model 7505 or equivalent.
    
          5.3.5    Aluminum foil - Heavy duty.
    
          5.3.6    Metal can - To recover used silica gel.
    
    5.4   Analysis
    
          5.4.1    Glass Soxhlet extractors - 40 mm ID complete with 45/50
                   S condenser, 24/40 $ 250 ml round-bottom flask, heating
                   mantle for 250 ml flask, and power transformer.
    
          5.4.2    Teflon FEP wash bottle - Two, 500 ml, Nalgene No. 0023A59
                   or equivalent.
    
          5.4.3    Separatory funnel - 1,000 ml with TFE® stopcock.
    
          5.4.4    Kuderna-Danish concentrators - 500 ml.
    
          5.4.5    Steam bath.
    
          5.4.6    Separatory funnel - 50 ml with TFE® stopcock.
    
          5.4.7    Volumetric flask - 25.0 ml, glass.
    
          5.4.8    Volumetric flask - 5.0 ml, glass.
    
          5.4.9    Culture tubes - 13 x 100 mm, glass with TFE®-lined screw
                   caps.
    
          5.4.10   Pipette - 5.0 ml glass.
    
          5.4.11   Teflon®-glass syringe - 1 ml, Hamilton 1001 TLL or
                   equivalent with Teflon® needle.
    
          5.4.12   Syringe - 10 (Jl, Hamilton 701N or equivalent.
                                 C-12
    

    -------
    5.4.13   Disposable glass pipettes with bulbs - To aid transfer
             of the extracts.
    
    5.4.14   Gas chromatography/mass spectrometer system.
    
             5.4.14.1  Gas chromatograph - An analytical system com-
                       plete with a temperature programmable gas chro-
                       matograph and all required accessories includ-
                       ing syringes, analytical columns, and gases.
                       The injection port must be designed for on-
                       coluran injection when using capillary columns
                       or packed columns.  Other capillary injection
                       techniques (split, splitless, "Grob," etc.)
                       may be used provided the performance specifi-
                       cations stated in Section 7.1 are met.
    
             5.4.14.2  Capillary GC column - A 12-20 m long x 0.25 mm
                       ID fused silica column with a 0.25 (J™ thick
                       DB-5 bonded silicone liquid phase (J&W Scien-
                       tific) is recommended.  Alternate liquid phases
                       may include OV-101, SP-2100, Apiezon L, Dexsil
                       300, or other liquid phases which meet the per-
                       formance specifications stated in Section 7.1.
    
             5.4.14.3  Packed GC column - A 180 cm x 0.2 cm ID glass
                       column packed with 3% SP-2250 on 100/120 mesh
                       Supelcoport or equivalent is recommended.
                       Other liquid phases which meet the performance
                       specifications stated in Section 7.1 may be
                       substituted.
    
             5.4.14.4  Mass spectrometer - Must be capable of scanning
                       from 150 to 550 daltons every 1.5 sec or less,
                       collecting at least five spectra per chromato-
                       graphic peak, utilizing a 70-eV (nominal) elec-
                       tron energy in the electron impact ionizaton
                       mode and producing a mass spectrum which meets
                       all the criteria in Table 2 when 50 ng of deca-
                       fluorotriphenyl phosphine [DFTPP, bis(perfluoro-
                       phenyl)phenyl phosphine] is injected through
                       the GC inlet.  Any GC-to-MS interface that
                       gives acceptable calibration points at 10 ng
                       per injection for each PCB isomer in the cali-
                       bration standard and achieves all acceptable
                       performance criteria (Section 10) may be used.
                       Direct coupling of the fused silica column to
                       the MS is recommended.  Alternatively, GC to
                       MS interfaces constructed of all glass or glass-
                       lined materials are recommended.  Glass can be
                       deactivated by silanizing with dichlorodimethyl-
                       silane.
                           C-13
    

    -------
       TABLE 2.  DFTPP KEY IONS AND ION ABUNDANCE CRITERIA
    
    Mass                           Ion abundance criteria
    
    
    197                            Less than 1% of mass 198
    198                            100% relative abundance
    199                            5-9% of mass 198
    
    275                            10-30% of mass 198
    
    365                            Greater than 1% of mass 198
    
    441                            Present, but less than mass 443
    442                            Greater than 40% of mass 198
    443                            17-23% of mass 442
                              C-14
    

    -------
                         5.4.14.5  A computer system that allows  the  continuous
                                   acquisition and storage on machine-readable
                                   media of all mass spectra  obtained throughout
                                   the duration of the  chromatographic program
                                   must be interfaced to  the  mass spectrometer.
                                   The data system must have  the  capability of
                                   integrating the abundances of  the  selected
                                   ions between specified limits  and  relating
                                   integrated abundances  to concentrations  using
                                   the calibration procedures described in  this
                                   method.  The computer  must have software that
                                   allows searching any GC/MS data file for ions
                                   of a specific mass and plotting such ion abun-
                                   dances versus time or  scan number  to yield an
                                   extracted ion current  profile  (EICP).  Software
                                   must also be available that allows integrating
                                   the abundance in any EICP  between  specified
                                   time or scan number  limits.
    6.0   Reagents
          6.1   Sampling
    
                6.1.1
          6.2
                6.1.2
    
    
    
                6.1.3
    
    
                6.1.4
             Florisil - Floridin Company,  30/60 mesh,  Grade A.   The
             Florisil is cleaned by 8 hr Soxhlet extraction with hex-
             ane and then by drying for 8  hr in an oven atvl!0°C and
             is activated by heating to 650°C for 2 hr (not to  exceed
             3 hr) in a muffle furnace. After allowing to cool to
             near 110°C transfer the clean,  active Florisil to  a clean,
             hexane-washed glass jar and seal with a TFE®-lined lid.
             The Florisil should be stored at 110°C until taken to
             the field for use.  Florisil  that has been stored  more
             than 1 month must be reactivated before use.
    
             Glass wool - Cleaned by thorough rinsing with hexane,
             dried in a 110°C oven, and stored in a hexane-washed
             glass jar with TFE®-lined screw cap.
    
             Water - Deionized, then glass-distilled,  and stored in
             hexane-rinsed glass containers with TFE®-lined screw caps.
    
             Silica gel - Indicating type, 6-16 mesh.   If previously
             used, dry at 175°C for 2 hr.   New silica gel may be used
             as received.
    6.1.5    Crushed ice.
    
    Solvents - All solvents must be pesticide residue analysis grade.
    New lots should be checked for purity by concentrating an aliquot
    by at least as much as is used in the procedure.
                                       C-15
    

    -------
    6.3   Calibration standard congeners - Standards of the PCB congeners
          listed in Table 3 are available from Ultra Scientific,  Hope,
          Rhode Island;  or Analabs,  North Haven,  Connecticut.
    
    6.4   Calibration standard stock solutions -  Primary dilutions of each
          of the individual PCBs listed in Table  3 are prepared by weighing
          approximately 1-10 mg of material within 1% precision.   The PCB
          is then dissolved and diluted to 1.0 ml with hexane.   The concen-
          tration is calculated in mg/ml.  The primary dilutions  are stored
          at 4°C in screw-cap vials  with Teflon cap liners.  The  meniscus
          is marked on the vial wall to monitor solvent evaporation.  Pri-
          mary dilutions are stable  indefinitely  if the seals are maintained.
          The validity of primary and secondary dilutions must be monitored
          on a quarterly basis by analyzing four  quality control  check sam-
          ples (see Section 14.2).
    
    6.5   Working calibration standards - Working calibration standards are
          prepared that are similar  in PCB composition and concentration to
          the samples by mixing and diluting the  individual standard stock
          solutions.  Example calibration solutions are shown in Table 3.
          The mixture is diluted to  volume with pesticide residue analysis
          quality hexane.  The concentration is calculated in ng/ml as the
          individual PCBs.  Dilutions are stored  at 4°C in narrow-mouth,
          screw-cap vials with Teflon cap liners.  The meniscus is marked
          on the vial wall to monitor solvent evaporation.  These secondary
          dilutions can be stored indefinitely if the seals are maintained.
          These solutions are designated "CSxxx," where the xxx is used to
          encode the nominal concentration in ng/ml.
    
    6.6   Alternatively, certified stock solutions similar to those listed
          in Table 3 may be available from a supplier, in lieu of the pro-
          cedures described in Section 6.4.
    
    6.7   DFTPP standard - A 50 ng/(Jl solution of DFTPP is prepared in ace-
          tone or another appropriate solvent.
    
    6.8   Internal standard stock solution - The  four 13C-labeled PCBs
          listed in Table 4 may be available from a supplier as a certi-
          fied solution.  This solution may be used as received or diluted
          further.
    
    6.9   Solution stability - The calibration standard, surrogate and DFTPP
          solutions should be checked frequently  for stability.  These solu-
          tions should be replaced after 6 months, or sooner if comparison
          with quality control check samples indicates compound degradation
          or concentration change.
    
    6.10  Quality control check samples will be supplied by the Agency.
                                 C-16
    

    -------
     TABLE 3.   CONCENTRATIONS OF CONGENERS IN PCS CALIBRATION STANDARDS (ng/ml)3
    
    Homo log
    1
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    4
    1
    4
    8
    10
    Congener
    no.
    1
    3
    7
    30
    50
    97
    143
    183
    202
    207
    209
    210 (IS)
    211 (RS)
    212 (RS)
    213 (RS)
    214 (RS)
    CS1000
    1,040
    1,000
    1,040
    1,040
    1,520
    1,740
    1,920
    2,600
    4,640
    5,060
    4,240
    255
    104
    257
    407
    502
    CS100
    104
    100
    104
    104
    152
    174
    192
    260
    464
    506
    424
    255
    104
    257
    407
    502
    CS050
    52
    50
    52
    52
    76
    87
    96
    130
    232
    253
    212
    255
    104
    257
    407
    502
    CS010
    10
    10
    10
    10
    15
    17
    19
    26
    46 .
    51
    42
    255
    104
    257
    407
    502
    
    a  Concentrations given as examples only.
                                      C-17
    

    -------
        TABLE 4.   COMPOSITION OF INTERNAL STANDARD SPIKING SOLUTION (SS100)
                            CONTAINING 13C-LABELED PCBsa
    
    Congener
    no.
    211
    212
    213
    214
    Compound
    (I1 ,2' ,3' ,4' ,5' ,6'-13C6)4-chlorobiphenyl
    (13Ci2)3>3' ,4,4'-tetrachlorobiphenyl
    (13C12)2,2' ,3,3' ,5,5' ,6,6'-octachlorobiphenyl
    (13Ci2)decachIorobiphenyl
    Concentration
    (Mg/ml)
    104
    257
    395
    502
    
    a  Concentrations given as examples only.
                                       C-18
    

    -------
    7.0   Calibration
    
          Maintain a laboratory log of all calibrations.
    
          7.1   Sampling train
    
                7.1.1    Probe nozzle - Using a micrometer, the inside diameter
                         of the nozzle is measured to the nearest 0.025 mm (0.001
                         in.).  Three separate measurements are made using differ-
                         ent diameters each time and obtain the average of the
                         measurements.  The difference between the high and low
                         numbers must not exceed 0.1 mm (0.004 in.).
    
                         When nozzles become nicked, dented, or corroded, they
                         must be reshaped, sharpened, and recalibrated before use.
    
                         Each nozzle must be permanently and uniquely identified.
    
                7.1.2    Pitot tube - The pitot tube must be calibrated according
                         to the procedure outlined in Method 2.2
    
                7.1.3    Dry gas meter and orifice meter - Both meters must be
                         calibrated according to the procedure outlined in APTD-
                         0581.3  When diaphragm pumps with bypass valves are used,
                         proper metering system design is checked by calibrating
                         the dry gas meter at an additional flow rate of 0.0057
                         m3/min (0.2 cfm) with the bypass valve fully opened and
                         then with it fully closed.  If there is more than ±2%
                         difference in flow rates when compared to the fully
                         closed position of the bypass valve, the system is not
                         designed properly and must be corrected.
    
                7.1.4    Probe heater calibration - The probe heating system must
                         be calibrated according to the procedure contained in
                         APTD-0581.3
    
                7.1.5    Temperature gauges - Dial and liquid filled bulb thermom-
                         eters are calibrated against mercury-in-glass thermometers,
                         Thermocouples should be calibrated in constant tempera-
                         ture baths.
    
          7.2   The gas chromatograph must meet the minimum operating parameters
                shown in Tables 5 and 6, daily.  If all of the criteria are not
                met, the analyst must adjust conditions and repeat the test until
                all criteria are met.
    
          7.3   The mass spectrometer must meet the minimum operating parameters
                shown in Tables 2, 7, and 8, daily.  If all criteria are not met,
                the analyst must retune the spectrometer and repeat the test un-
                til all conditions are met.
                                       C-19
    

    -------
    TABLE 5.  OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM
            Parameter
          Recommended
        Tolerance
    Gas chromatograph
    Column
    
    Liquid phase
    
    Liquid phase thickness
    Carrier gas
    Carrier gas velocity
    Injector
    Injector temperature
    Injection volume
    Initial column temperature
    Column temperature program
    Finnigan 9610
    15 ra x 0.255 mm ID
    Fused silica
    DB-5 (J&W)
    0.25 pm
    Helium
    /i-   /   b
    45 cm/sec
    On-column (J&W)C
    Optimum performance
    1.0 Ml°
    70°C (2 min)d
    70°-325°C at 10°C/min£
    Other
    Other
    
    Other nonpolar
    or semipolar
    < 1 pm
    Hydrogen
    Optimum performance
    Other
    Optimum performance
    Other
    Other
    Other
    Separator
    Transfer line temperature
    Tailing factorh
    Peak width1
    None1"
    280°C
    0.7-1.5
    7-10 sec
    Glass jet or othe
    Optimum**
    0.4-3
    < 15 sec
    
    a  Substitutions permitted with any common apparatus or technique provided
       performance criteria are met.
    b  Measured by injection of air or methane at 270°C oven temperature.
    c  For on-column injection, manufacturer's instructions should be followed
       regarding injection technique.    .         ,
    d  With on-column injection, initial temperature equals boiling point of the
       solvent; in this instance, hexane.
    e  Cj2Cli6 elutes at 270°C.  Programming above this temperature ensures a
       clean column and lower background on subsequent runs.
    f  Fused silica columns may be routed directly into the ion source to prevent
       separator discrimination and losses.
    g  High enough to elute all PCBs, but not high enough to degrade the column
       if routed through the transfer line.
    h  Tailing factor is width of front half of peak at 10% height divided by
       width of back half of peak at 10% height for single PCB congeners in solu-
       tion CSxxx.
    i  Peak width at 10% height for a single PCB congener is CSxxx.
                                        C-20
    

    -------
     TABLE 6.  OPERATING PARAMETERS FOR PACKED COLUMN GAS CHROMATOGRAPHY SYSTEM
          Parameter
       Recommended
       Tolerance
    Gas chromatograph
    Column
    Finnigan 9610
    180 cm x 0.2 cm ID
    Other3
    Other
    Column packing
    
    
    Carrier gas
    
    Carrier gas flow rate
    
    Injector
    
    Injector temperature
    
    Injection volume
    
    Initial column temperature
    
    Column temperature program
    
    Separator
    
    Transfer line temperature
    
    Tailing factor
    
    Peak widthd
    glass
    
    3% SP-2250 on  100/
    120 mesh Supelcoport
    
    Helium
    
    30 ml/min
    
    On-column
    
    250°C
    
    1.0 pi
    
    150°C, 4 min
    
    150°-260°C at  8°/min
    
    Glass jet
    
    280°C
    
    ,0.7-1.5
    
    10-20 sec
    Other nonpolar
    or semipolar
    
    Hydrogen
    
    Optimum performance
    
    Other
    
    Optimum
    
    ^ 5 Ml
    
    Other
    
    Other
    
    Other
    
    Optimum3
    
    0.4-3
    
    < 30 sec
    a  Substitutions permitted if performance criteria are met.
    
    b  High enough to elute all PCBs.
    
    c  Tailing factor is width of front half of peak at 10% height divided by
       width of back half of peak at 10% height for single PCB congeners in solu-
       tion CSxxx.
    
    d  Peak width at 10% height for a single PCB congener is CSxxx.
                                       C-21
    

    -------
       TABLE 7.   OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER SYSTEM
    
          Parameter                    Recommended                Tolerance
    Mass spectrometer
    Data system
    Scan range
    Scan time
    Resolution
    Ion source temperature
    Electron energy
    Trap current
    Multiplier voltage
    Preamplifier sensitivity
    Finnigan 4023
    Incos 2400
    95-550
    1 sec
    Unit
    280°C
    70 eV
    0.2 mA
    -1,600 V
    10"6 A/V
    Other3
    Other
    Other
    Otherb
    Optimum performance
    200°-300°C
    Optimum performance
    Optimum performance
    Optimum performance
    Set for desired
    working range
    
    a  Substitutions permitted if performance criteria are met.
    
    b  Greater than five data points over a GC peak is a minimum.
    
    c  Filaments should be shut off during solvent elution to improve instrument
       stability and prolong filament life, especially if no separator is used.
                                     C-22
    

    -------
     TABLE 8.   OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER SYSTEM
    
    Parameter
    Mass spectrometer
    Data system
    Scan range
    Scan mode
    Cycle time
    Resolution
    Ion source temperature
    Electron energy
    Emission current
    Filament current
    Multiplier
    Recommended
    Finnigan MAT 311A
    Incos 2400
    98-550
    Exponential
    1.2 sec
    1,000
    280°C
    70 eV
    1-2 mA
    Optimum
    -1,600 V
    Tolerance
    Other3
    Other
    Other
    Other
    Otherb
    > 500
    250-300°
    70 eV
    Optimum
    Optimum
    Optimum
    
    a  Substitutions permitted if performance criteria are met.
    
    b  Greater than five data pointy over a GC peak is a minimum.
    
    c  Filaments should be shut off during solvent elution to improve instrument
       stability and prolong filament life, especially if no separator is used.
                                       C-23
    

    -------
    7.4   The PCB response factors (RF ) must be determined using Equation
          7-1 for the analyte homologs?
    
                          A  x M.
                    RF  = v£	^                                Eq.  7-1
                      p   A.  x M                                  ^
                      r    is    p
    
          where    RF  = response factor of a given PCB isomer
    
                    A  = area of the characteristic ion for the PCB congener
                     "     peak
    
                    M  = mass of PCB congener injected (nanograms)
    
                   A.  = area of the characteristic ion for the internal
                           standard peak
    
                   M.  = mass of internal standard injected (nanograms)
                    X. S
    
          If specific congeners are known to be present and if standards
          are available, selected RF values may be employed.   For general
          samples, solutions CSxxx and SSxxx or a mixture (Tables 3 and 4)
          may be used as the response factor solution.  The PCB-surrogate
          pairs to be used in the RF calculation are listed in Table 9.
    
          Generally, only the primary ions of both the analyte and surrogate
          are used to determine the RF values.  If alternate ions are to be
          used in the quantitation, the RF must be determined using that
          characteristic ion.
    
          The RF value must be determined in a manner to assure ±20% accu-
          racy and precision.  For instruments with good day-to-day preci-
          sion, a running mean (RF) based on seven values determined once
          each day may be appropriate.  Other options include, but are not
          limited to, triplicate determinations of a single concentration
          spaced throughout a day or determination of the RF at three dif-
          ferent levels to establish a working curve.
    
          If replicate RF values differ by greater than ±10% RSD, the system
          performance should be monitored closely.  If the RSD is greater
          than ±20%, the data set must be considered invalid and the RF re-
          determined before further analyses are done.
    
    7.5   If the GC/EIMS system has not been demonstrated to yield a linear
          response or if the analyte concentrations are more than one order
          of magnitude different from those in the RF solution, a calibra-
          tion curve must be prepared.  If the analyte and RF solution con-
          centrations differ by more than one order of magnitude, a calibra-
          tion curve should be prepared.  A calibration curve should be
          established with triplicate determinations at three or more con-
          centrations bracketing the analyte levels.
                                 C-24
    

    -------
                         TABLE 9.   PAIRINGS OF ANALYTE,  CALIBRATION.  AND SURROGATE COMPOUNDS
    
    
    
    Analyte
    Congener
    no . Compound
    
    
    
    
    
    
    Calibration standard
    Congener
    no.
    1 2-Ci2H9Cl 1
    2,3 3- and 4-C12H9Cl 3
    4-15 C12HaCl2 7
    
    
    
    
    
    
    
    o
    10
    i c. on f* u r*~\
    ID™ Jy LI 2X170-1.3
    40-81 C12H6C14
    82-127 C12H5C15
    128-169 C12H4C16
    •f T/\ "f f\ O f* TT f* 1
    i/U-iyo Li2n3Lil7
    monc r> u PI
    -/Ub Li2n2Llg
    206—208 Ci2HCl9
    209 Ci2Clio
    
    
    30
    50
    97
    143
    183
    202
    207
    209
    
    
    2
    4
    2
    2
    2
    2
    2
    ?
    ?,
    7
    C
    
    
    ,4
    ,4,
    ,2'
    ,2'
    ,2'
    ,2'
    ,2'
    ?'
    12C
    
    
    
    6
    ,4
    ,3
    ,3
    ,3
    ,3
    3
    ll
    
    
    Compound
    
    
    ,6
    ',4,
    ,4,5
    ',4,
    3'
    
    0
    
    
    
    
    
    5
    ,6'
    4' ,5', 6
    5, 5', 6, 6'
    4, 4', 5, 6, 6'
    
    
    
    
    
    
    
    Surrogate
    Congener
    no.
    211
    211
    211
    212
    212
    212
    212
    213
    213
    213
    214
    
    
    Compound
    J3C6-4
    13Cg-4
    13C12-3,3'
    13Ci2-3,3'
    13Ci2-3,3'
    1 3p _ o o t
    13C12-2,2'
    13C12-2,2'
    13C12-2,2'
    13Ci2Cll0
    
    
    
    ,4,4'
    ,4,4'
    ,4,4'
    ,4,4'
    ,3, 3', 5, 5'
    ,3, 3', 5, 5'
    ,3,3', 5, 5'
    
    
    
    
    
    
    
    
    ,6,6'
    ,6,6'
    ,6,6'
    
    
    
    a  Ballschmiter numbering system, see Table 1.
    

    -------
          7.6   The relative retention time (RRT)  windows for the 10 horaologs and
                surrogates must be determined.   If all congeners are not available,
                a mixture of available congeners or an Aroclor mixture (e.g.,
                1016/1254/1260) may be used to  estimate the windows.  The windows
                must be set wider than observed if all isomers are not determined.
                Typical RRT windows for one column are listed in Table 10.   The
                windows may differ substantially if other GC parameters are used.
    
    
    8.0   Sample Collection, Handling, and Preservation
    
          The sampling shall be conducted by competent personnel experienced with
          this test procedure and cognizant of  the constraints of the anaytical
          techniques for PCBs, particularly contamination problems.
    
          8.1   Stack sampling1
    
                8.1.1    Pretest preparation -  All train components shall be main-
                         tained and calibrated  according to the procedure de-
                         scribed in APTD-0581,3 unless otherwise specified herein.
                         This should be done in the laboratory prior to sampling.
    
                         8.1.1.1  Cleaning glassware - All glass parts of the
                                  train upstream of and including the adsorbent
                                  tube and impingers,  should be cleaned as de-
                                  scribed in Section 3.1.1.  Special care should
                                  be devoted to the removal of residual silicone
                                  grease sealants, on ground glass connections of
                                  used glassware.   These grease residues should
                                  be removed by soaking several hours in a chromic
                                  acid cleaning solution prior to routine cleaning
                                  as described  above.
    
                         8.1.1.2  Solid adsorbent tube - 7.5 g of Florisil acti-
                                  vated within  the last 30 days and still warm
                                  from storage  in a 110°C oven, is weighed into
                                  the adsorbent tube (prerinsed with hexane) with
                                  a glass wool  plug in the downstream end.   A
                                  second glass  wool plug is placed in the tube to
                                  hold the sorbent in the tube.  Both ends of the
                                  tube are capped with ground glass caps.  These
                                  caps should not be removed until the tube is
                                  fitted to the train immediately prior to sampling.
    
                8.1.2    Preliminary determinations - The sampling site and the
                         minimum number of sampling points are selected according
                         to Method I2 or as specified by the Agency.  The stack
                         pressure, temperature, and the range of velocity heads
                         are determined using Method 22 and moisture content using
                         Approximation Method 42 or its alternatives for the pur-
                         pose of making isokinetic sampling rate calculations.
                         Estimates may be used.  However, final results must be
                         based on actual measurements made during the test.
    
                                       C-26
    

    -------
           TABLE 10.   RELATIVE RETENTION TIME (RRT)  RANGES OF PCB HOMOLOGS
                       VERSUS d6-3.3'.4,4'-TETRACHLOROBIPHENYL
    
    No. of
    PCB isomers
    homolog measured
    Monochloro
    Dichloro
    Trichloro
    Tetrachloro
    Pentachloro
    Hexachloro
    Heptachloro
    Octachloro
    Nonachloro
    Decachloro
    3
    10
    9
    16
    12
    13
    4
    6
    3
    1
    Observed range
    of RRTsa
    0.40-0.50
    0.52-0.69
    0.62-0.79
    0.72-1.01
    0.82-1.08
    0.93-1.20
    1.09-1.30
    1.19-1.36
    1.31-1.42
    1.44-1.45
    Congener
    no.
    1
    3
    7
    30
    50
    97
    143
    183
    202
    207
    209
    Observed
    RRTa
    0.43
    0.50
    0.58
    0.65
    0.75
    0.98
    1.05
    1.15
    1.19
    1.33
    1.44
    Projected
    range of
    RRTs
    0.35-0.55
    0.35-0.80
    0.35-0.10
    0.55-1.05
    0.80-1.10
    0.90-1.25
    1.05-1.35
    1.10-1.50
    1.25-1.50
    1.35-1.50
    
    a  The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were
       measured versus 3,3',4,4'-tetrachlorobiphenyl-de (internal standard) using
       a 15-m J&W DB-5 fused silica column with a temperature program of 110°C
       for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an on-
       column injector.  A Finnigan 4023 Incos quadrupole mass spectrometer oper-
       ating with a scan range of 95-550 daltons was used to detect each PCB
       congener.
    
    b  The projected relative retention windows account for overlap of eluting
       homologs and take into consideration differences in operating systems and
       lack of all possible 209 PCB congeners.
                                      C-27
    

    -------
             The molecular weight of the stack gases is determined
             using Method 3.2
    
             A nozzle size is selected based on the maximum velocity
             head so that isokinetic sampling can be maintained at a
             rate less than 0.75 cfm.   It is not necessary to change
             the nozzle size in order to maintain isokinetic sampling
             rates.  During the run, the nozzle size must not be
             changed.
    
             A suitable probe length is selected such that all traverse
             points can be sampled.   Sampling from opposite sides for
             large stacks may be considered to reduce the length of
             probes.
    
             A sampling time is selected appropriate for total method
             sensitivity and the PCB concentration anticipated.   Sam-
             pling times should generally fall within a range of 2 to
             4 hr.
    
             A buzzer-timer should be incorporated in the control box
             (see Figure 1) to alarm the operator to move the probe to
             the next sampling point.
    
    8.1.3    Preparation of collection train - During preparation and
             assembly of the sampling train, all train openings must
             be covered until just prior to assembly or until sampling
             is about to begin.  Immediately prior to assembly,  all
             parts of the train upstream of the adsorbent tube are
             rinsed with hexane.  The probe is marked with heat resis-
             tant tape or by some other method at points indicating
             the proper distance into the stack or duct for each sam-
             pling point.
    
             200 ml of water is placed in each of the first two impin-
             gers, and the third impinger left empty.  CAUTION:   Sealant
             greases must not be used in assembling the train.  If the
             preliminary moisture determination shows that the stack
             gases are saturated or supersaturated, one or two addi-
             tional empty impingers should be added to the train be-
             tween the third impinger and the Florisil tube.  See
             Section 5.1.5.  Approximately 200 to 300 g or more, if
             necessary, of silica gel is placed in the last impinger.
             Each impinger (stem included) is weighed and the weights
             recorded to the nearest 0.1 g on the impingers and on
             the data sheet.
    
             Unless otherwise specified by the Agency, a temperature
             probe is attached to the metal sheath of the sampling
             probe so that the sensor is at least 2.5 cm behind the
             nozzle and pitot tube and does not touch any metal.
                           C-28
    

    -------
             The train is assembled as shown in Figure 1.   Through all
             parts of this method use of sealant greases such as stop-
             cock grease to seal ground glass joints must be avoided.
    
             Crushed ice is placed around the impingers.
    
    8.1.4    Leak check procedure - After the sampling train has been
             assembled, the probe heating system(s) is turned on and
             set (if applicable) to reach a temperature sufficient to
             avoid condensation in the probe.  Time is allowed for the
             temperature to stabilize.  The train is leak checked at
             the sampling site by plugging the nozzle and pulling a
             380 mm Hg (15 in. Hg) vacuum.  A leakage rate in excess
             of 4% of the average sampling rate or 0.0057 m3/min
             (0.02 cfm) whichever is less, is unacceptable.
    
             The following leak check instruction for the sampling
             train described in APTD-05813 may be helpful.  The pump
             is started with bypass valve fully open and coarse adjust
             valve completely closed.  The coarse adjust valve is
             partially opened and the bypass valve slowly closed until
             380 mm Hg (15 in. Hg) vacuum is reached.  The direction
             of bypass valve must not be reversed.  This will cause
             water to back up into the probe.  If 380 mm Hg (15.in. Hg)
             is exceeded, either the leak check is conducted at this
             higher vacuum or the leak check is ended as described
             below and start over.
    
             When the leak check is completed, the plug is first slowly
             removed from the inlet to the probe and the vacuum pump
             is immediately turned off.  This prevents the water in
             the impingers from being forced backward into the probe.
    
             Leak checks, shall be conducted as described above prior
             to each test run and at the completion of each test run.
             If leaks are found to be in excess of the acceptable rate,
             the test will be considered invalid.  To reduce lost time
             due to leakage occurrences, it is recommended that leak
             checks be conducted between port changes.
    
    8.1.5    Train operation - During the sampling run, an isokinetic
             sampling rate within 10%, or as specified by the Agency,
             of true isokinetic shall be maintained.  During the run,
             the nozzle or any other part of the train in front of
             and including the Florisil tube must not be changed.
    
             For each run, the data required on the data sheets must
             be recorded.  An example is shown in Figure 4.  The dry
             gas meter readings are recorded at the beginning and end
             of each sampling time increment, when changes in flow
             rates are made, and when sampling is halted.   Other data
             point readings are taken at least once at each sample
             point during each time increment and whenever significant
    
                           C-29
    

    -------
                                                                             FIELD DATA
                             PLANT.
                             DATE._
                             SAMPLING LOCATION .
                             SAMPLE TYPE 	
                             RUN NUMBER.	
                             OPERATOR	
                                               PROBE LENGTH AND TYPE.
                                               NOZZLE I.D..,.. 	
                                               ASSUMED MOISTURE. °,	
                             AMBIENT TEMPERATURE
                             BAROMETRIC PRESSURE .
                             STATIC PRESSURE. (Ps)_
                             FILTER NUMBER (s)	
                                               SAMPLE BOX NUMBER	
                                               METER BOX NUMBER	
                                               METER AHf,	
                                               C FACTOR	
                                               PROBE HEATER SETTING	
                                               HEATER BOX SETIING_	
                                               REFERENCE Ap	
                                                                      SCHEMATIC OF TRAVERSE POINT LAYOUT
                                                               READ AND RECORD ALL DATA EVERY.
                                          MINUTES
    TRAVERSE
    POINT
    NUMBER
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    \. CLOCK TIME
    ™fLING XnS™
    TIMt.mm N^
    	 	 	 __
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    GAS METER READING
    (Vm). It3
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    VELOCITY
    HEAD
    (&PS). in. H?0
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    ORIFICE PRESSURE
    DIFFERENTIAL
    (AH), in. HjOl
    DESIRED
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    ACTUAL
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    STACK
    TEMPERATURE
    ."F
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    DRY GAS METER
    TEMPERATURE
    INLET
    
    -------
    changes (20% variation in velocity head readings) neces-
    sitate additional adjustments in flow rate.
    
    The portholes are cleaned prior to the test run to mini-
    mize change of sampling deposited material.  To begin
    sampling, the nozzle cap is removed, the probe heater
    operational and temperature up, and the pitot tube and
    probe positions are verified (if applicable).   The nozzle
    is positioned at the first traverse point with the tip
    pointing directly into the gas stream.  The pump is
    started and the flow adjusted to isokinetic conditions.
    Nomographs are available for sampling trains using type
    S pitot tubes with 0.85 ± 0.02 coefficients (C ), and
    when sampling in air or a stack gas with equivalent
    density (molecular weight, M,, equal to 29 ± 4), which
    aid in the rapid adjustment of the isokinetic sampling
    rate without excessive computations.  If C  and M, are
    outside the above stated ranges, the nomograph cannot be
    used unless appropriate steps are taken to compensate for
    the deviations.
    
    When the stack is under significant negative pressure
    (height of impinger stem), the coarse adjust valve must
    be closed before inserting the probe into the stack to
    avoid water backing into the probe.  If necessary, the
    pump may be turned on with the coarse valve closed.
    
    When the probe is in position, the openings around the
    probe and porthole must be blocked off to prevent un-
    representative dilution of the gas stream.
    
    The stack cross section is traversed, as required by
    Method I2 or as specified by the Agency.  To minimize
    chance of extracting deposited material, the probe nozzle
    should not bump into the stack walls when sampling near
    the walls or when removing or inserting the probe through
    the portholes.
    
    During the test run, periodic adjustments are made to
    keep the probe temperature at the proper value.  More
    ice and, if necessary, salt is added to the ice bath to
    maintain a temperature of less than 20°C (68°F) at the
    impinger/silica gel outlet, to avoid excessive moisture
    losses.  Also, the level and zero of the manometer should
    be periodically checked.
    
    If the pressure drop across the train becomes high enough
    to make isokinetic sampling difficult to maintain, the
    test run should be terminated.  Under no circumstances
    should the train be disassembled during the test run to
    determine and correct causes of excessive pressure drops.
                  C-31
    

    -------
                   At the end of the sample run, the pump is turned off,  the
                   probe and nozzle removed from the stack, and the final
                   dry gas meter reading recorded.   A leak check is performed,
                   with acceptability of the test run based on the same cri-
                   teria as in Section 8.1.4.   The  percent isokinetic is
                   calculated (see calculation section) to determine whether
                   another test run should be made.  If there is difficulty
                   in maintaining isokinetic rates  due to source conditions,
                   the Agency should be consulted for possible variance on
                   the isokinetic rates.
    
          8.1.6    Blank train - For each series of test runs, a blank train
                   is set up in a manner identical  to that described above,
                   but with the nozzle capped with  aluminum foil and the
                   exit end of the last impinger capped with a ground glass
                   cap.  The train is allowed to remain assembled for a
                   period equivalent to one test run.  The blank sample is
                   recovered as described in Section 8.3.
    
    8.2   Static air sampling3 - The sampling procedure for static air is
          identical to that described in Section 8.1 with the following ex-
          ceptions:  (a) impingers and a beatable probe are not required
          prior to the adsorbent tube; and (b) the  PCB concentrations may
          dictate a longer or shorter sampling time.
    
          The selection of sampling time and rate should be based on the
          approximate levels of PCB residues expected in the sample.  The
          sampling rate should not exceed 14 liter/min and may typically
          fall in the range of 5 to 10 liter/min.  Sampling times should  be
          more than 20 min but should not exceed 4  hr.
    
    8.3   Sample recovery - Proper cleanup procedure begins as soon as the
          probe is removed from the stack at the end of the sampling period.
    
          When the probe can be safely handled, all external particulate
          matter near the tip of the probe nozzle is wiped off.  The probe
          is removed from the train and both ends closed off with aluminum
          foil.  The inlet to the train is capped off with a ground glass
          cap.
    
          The probe and impinger assembly are transfered to the cleanup area.
          This* area should be clean and protected from the wind so that the
          chances of contaminating or losing the sample will be minimized.
    
          The train is inspected prior to and during disassembly and any
          abnormal conditions noted.  The samples are treated as follows:
    
          8.3.1    Adsorbent tube - The Florisil tube is removed from the
                   train and capped with ground glass caps.
    
          8.3.2    Sample Container No. 1 - The first three impingers are
                   removed.  The outside of each impinger is wiped off to
                   remove excessive water and other debris.  The impingers
    
                                 C-32
    

    -------
                         are weighed (stem included),  and  the  weight  recorded  on
                         a data sheet.   The contents  are poured  directly into
                         Container No.  1.
    
                8.3.3    Sample Container  No.  2 -  Each of  the  first three impingers
                         are rinsed sequentially with 30-ml  acetone and then with
                         30-ml hexane,  and the rinses put  into Container No. 2.
                         Material deposited in the probe is  quantitatively recov-
                         ered using 100-ml acetone and then  100-ml hexane and
                         these rinses added to Container No.  2.
    
                8.3.4    Silica gel container  - The last impinger is  removed,  and
                         the outside wiped to  remove  excessive water  and other
                         debris.   It is weighed (stem included), and  the weight
                         recorded on the data  sheet.   The  contents are transferred
                         to the used silica gel can.
    
          8.4   Sample preservation - Samples  should  be stored in the dark at  4°C.
                Storage times in excess of 4 weeks are not recommended.
    
    
    9.0  Sample Preparation1
    
          9.1   Extraction
    
                9.1.1    Adsorbent tube -  The  entire  contents  of the  adsorbent
                         tube are expelled directly onto a glass wool plug in  the
                         sample holder of a Soxhlet extractor.  Although no extrac-
                         tion thimble is required, a  glass thimble with a coarse-
                         fritted bottom may be used.
    
                         The tube is rinsed with 5-ml acetone and then with 15-ml
                         hexane and these rinses put into  the extractor.  The  ex-
                         traction apparatus is assembled and the adsorbent ex-
                         tracted with 170-ml hexane for at least 4 hr.  The ex-
                         tractor should cycle  10 to 14 times per hour.  After
                         allowing the extraction apparatus to cool to ambient
                         temperature, the extract is  transferred into a Kuderna-
                         Danish evaporator.
    
                         The extract is evaporated to about  5 ml on a steam bath
                         and the evaporator allowed to cool  to ambient temperature
                         before disassembly.  The extract  is transferred to a  50-ml
                         separatory funnel and the funnel  set aside.
    
                9.1.2    Sample Container No.  1 -  The aqueous sample  is transferred
                         to a 1,000-ml separatory funnel.   The container is rinsed
                         with 20-ml acetone and then with  two 20-ml portions of
                         hexane,  adding the rinses to the  separatory  funnel.
    
                         The sample is extracted with three  100  ml portions of
                         hexane and the sequential extracts  transferred to a
                         Kuderna-Danish evaporator.
    
                                       C-33
    

    -------
                   The extract is concentrated to about 5 ml and allowed to
                   cool to ambient temperature before disassembly.   The ex-
                   tract is filtered through a micro column of anhydrous
                   sodium sulfate into a 50-ml separatory funnel containing
                   the corresponding Florisil extract from Section 9.1.1.
                   The micro column is prepared by placing a small plug of
                   glass wool in the bottom of the large portion of a dis-
                   posable pipette and then adding anhydrous sodium sulfate
                   until the tube is about half full.
    
          9.1.3    Sample Container No.  2 - The organic solution is trans-
                   ferred into a 1,000-ml separatory funnel.  The container
                   is rinsed with two '20 ml portions of hexane and the rinses
                   added to the separatory funnel.  The sample is washed with
                   three 100 ml portions of water.  The aqueous layer is
                   discarded and the organic layer transferred to a Kuderna-
                   Danish evaporator.
    
                   The extract is concentrated to about 5 ml and allowed to
                   cool to ambient temperature before disassembly.   The ex-
                   tract is filtered through a micro column of anhydrous
                   sodium sulfate into the 50-ml separatory funnel contain-
                   ing the corresponding Florisil and impinger extracts
                   (Section 9.1.2).
    
    9.2   Cleanup - Two tested cleanup techniques are described below.4  De-
          pending upon the complexity of the sample, one or both of the tech-
          niques may be required to fractionate the PCBs from interferences.
          If the sample extract is colored, the Florisil column cleanup may
          be indicated.
    
          9.2.1    Acid cleanup
    
                   9.2.1.1  Add 5 ml of concentrated sulfuric acid to the
                            separatory funnel containing the sample extract
                            and shake for 1 min.
    
                   9.2.1.2  Allow the phases to separate, transfer the
                            sample (upper phase) with three 1 to 2 ml
                            solvent rinses to Kuderna-Danish evaporator
                            and concentrate to an appropriate volume.
    
                   9.2.1.3  Analyze as described in Section 10.0.
    
                   9.2.1.4  If the sample is highly contaminated, a second
                            or third acid cleanup may be employed.
    
          9.2.2    Florisil column cleanup
    
                   9.2.2.1  Variations among batches of Florisil may affect
                            the elution volume of the various PCBs.  For
                            this reason, the volume of solvent required to
                                 C-34
    

    -------
                                  completely elute  all  of  the  PCBs must be  veri-
                                  fied by the analyst.   The  weight of Florisil
                                  can then be adjusted  accordingly.
    
                         9.2.2.2  Place a 20-g charge of Florisil, activated  over-
                                  night at 130°C, into  a Chromaflex  column.   Settle
                                  the Florisil by tapping  the  column.  Add  about
                                  1  cm of anhydrous sodium sulfate to the top of
                                  the Florisil.  Pre-elute the column with  70-80
                                  ml of hexane.  Just before the  exposure of  the
                                  sodium sulfate layer  to  air,  stop  the flow.
                                  Discard the eluate.
    
                         9.2.2.3  Add the sample extract to  the column.  Add  225
                                  ml of hexane to the column.   Carefully wash
                                  down the inner wall of the column  with a  small
                                  amount of the hexane  prior to adding the  total
                                  volume.  Discard  the  first 25 ml.
    
                         9.2.2.A  Collect 200 ml of hexane eluate in a Kuderna-
                                  Danish flask.  All of the  PCBs  should be  in
                                  this fraction. Concentrate  to  an  appropriate
                                  volume.
    
                         9.2.2.5  Analyze the sample as described in Section  10.0.
    
    
    10.0  Gas Chromatographic/Electron Impact Mass  Spectrometric  Determination
    
          10.1  Internal standard addition - Pipet  an appropriate volume of inter-
                nal standard solution SSxxx into the sample.  The final concentra-
                tion of the internal standards must be  in  the  working range of the
                calibration and well above the matrix background. The internal
                standards are thoroughly mixed by mechanical agitation.
    
                Note:  The volume measurement of the spiking solution is critical
                to the overall method precision. The analyst  must exercise cau-
                tion that the volume is known ±1% or better.  Where  necessary,
                calibration of the pipet is recommended.
    
                Note:  This same solution is used as a  surrogate  standard solution
                in the protocols for products/product waste  and for  water.  In
                this protocol, the 13C-labeled PCBs are spiked after extraction,
                so are used as internal standards.
    
                Alternately, another internal standard  solution such as the dg-
                3,3',4,4'-tetrachlorobiphenyl used  in the  product/product waste
                and water protocols  may be used, if acceptable RF precision and
                accuracy are shown across the homolog range.
                                                 i
          10.2  Tables 2, and 5  through 8 summarize the recommended  operating con-
                ditions for analysis.  Figure 5 presents an  example  of a chromato-
                gram.
    
                                       C-35
    

    -------
               iet).o-i
                me
    o
    i
    LO
    en cu
    4-J
    ,H CO
    cfl 00
    C 0
    M r-l
    CU r-l
    4J 3
    C "I
    i- ^
    vO vO CO
    O 13 rH
    rH rH rH
    1 X X
    rH rH . C C
    XX CU CU
    c c x; x;
    (U CU P. p.
    Xi XI "r-t "H
    P. P. Xi X)
    •r( -H O O
    Xl XI ^~^ r-llfi.V'jH
    O O CU o
    M VJ J-l rH
    o o CD x;
    rH rH 00 O
    Xi Xi O cfl
    O O r-l r-*
    O O M rH 4-1
    C C 3 X cu
    O O W C H
    a s ^ cu i
    II rC
     rH H Xi -
    
    rH
    £>-,
    C
    CU
    x:
    P.
    •H
    Xi
    o
    ^i
    o
    rH
    Xi
    o
    o
    C
    o
    ^
    1
    CN
    
    
    
    018
    |
    
    X O P. CO
    C rH -H
    rH cu x; - xi co
    x xi o o
    C P, Cfl rJ
    CU -H r-l O
    Xi XI U rH
    P. O (U X!
    •H M H a
    XI O I cfl
    O rH VO -U
    MX:- c
    o a 
    
    ft ^ t
    T T
    L L I « II . ». - 11.
    
    rH
    x;
    o
    CO rH
    V-i X
    •u C
    a» cu
    H x:
    i ex
    •H
    ^J" XI
    « o
    -^J- ^
    - o
    - rH
    co x;
    •- o
    PO tfl
    rN
    CU
    pe
    
    W
    vD
    «
    m
    r.
    ^-
    A
    ^J-
    *
    CO
    
    
    
    
    
    
    
    
    tl
    
    rH
    rH C
    X cu
    C JC
    CU p.
    Xi -H rH
    P. X) X
    •H 0 C
    XI MO)
    o ox;
    r-l rH P.
    o x: -H
    rH O XI
    X! CO 0
    O C r-l
    CO 0 O
    4-1 K rH
    a i x!
    O 0
    1 ^O Cfl
    o
    vO v£> (1)
    Q
    vO iO
    rv r.
    iO <3"
    *. *%
    •O ••^"
    f. ft
    +. —
    ro ro
    
    CO CO
    ft ft
    *. —
    CN CN
    r. ft
    CN OJ
    
    
    
    ^N .200 K °
    •1
    CN
    
    
    
    II7U
    
    
    
    12."
    20:
    
    
    
    
    
    
    
    
    ,»„ _ ^Jl 	
    "
    ' 112592.
    
    CN
    1 — 1
    U
    CO
    rH
    1
    rH
    
    C
    CU
    x;
    P, ,—>
    •H CU
    f"1 4J
    O cfl
    M Ml
    0 0
    i — 1 M
    Xi r-l
    O 3
    cfl o)
    o ^
    CU
    Q
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    _H
    3 I'.'f'j SCA:I
    C!) 23:20 1 1 III i
           Figure 5.  Capillary gas chromatography/electron impact  ionization mass spectrometry (CGC/EIMS)
             chromatogram or the calibration standard  solution required for quantitatlon of PCBs by homolog.
             This chromatogram includes PCBs representative of each homolog, three carbon-13 labeled surrogates,
             and  the deuterated internal standard.   The concentration of all components and the CGC/EIMS
             parameters are presented  in 'Cables  3,  4,  5, and  7.
    

    -------
          10.3  While the highest available  chromatographic  resolution is  not a
                necessary objective of this  protocol,  good chromatographic per-
                formance is recommended.   With the high resolution of  CGC, the
                probability that the chromatographic peaks consist of  single
                compounds is higher than  with PGC.  Thus,  qualitative  and  quanti-
                tative data reduction should be more reliable.
    
          10.4  After performance of the  system has been certified for the day
                and all instrument conditions set according  to  Tables  2,  and 5
                through 8, inject an aliquot of the sample onto the GC column.
                If the response for any ion, including surrogates  and  internal
                standard, exceeds the working range of the system, dilute  the
                sample and reanalyze.  If the responses of surrogates, internal
                standard, or analytes are below the working  range, recheck the
                system performance.  If necessary, concentrate  the sample  and
                reanalyze.
    
          10.5  Record all data on a digital storage device  (magnetic  disk, tape,
                etc.) for qualitative and quantitative data  reduction  as  discussed
                below.
    11.0  Qualitative Identification
    
          11.1  Selected ion monitoring (SIM)  or limited mass scan (IMS)  data -
                The identification of a compound as a given PCB homolog requires
                that two criteria be met:
    
                11.1.1   (1) The peak must elute within the retention time window
                         set for that homolog (Section 7.6); and (2)  the  ratio of
                         two ions obtained by SIM (Table 11) or by IMS (Table 12)
                         must match the natural ratio within ±20%.  The analyst
                         must search the higher mass windows, in particular M+70,
                         to prevent misidentification of a PCB fragment ion clus-
                         ter as the parent.
    
                11.1.2   If one or the other of these criteria is not met, inter-
                         ferences may have affected the results and a reanalysis
                         using full scan EIMS conditions is recommended.
    
          11.2  Full scan data
    
                11.2.1   The peak must elute within the retention time windows
                         set for that homolog (as described in Section 7.6).
    
                11.2.2   The unknown spectrum must match that of an authentic PCB.
                         The intensity of the three largest ions in the molecular
                         cluster (two largest for monochlorobiphenyls) must match
                         the natural ratio within ±20%.  Frequent clusters with
                         proper intensity ratios must also be present.
    
                11.2.3   Alternatively, a spectral search may be used to  auto-
                         matically reduce the data.  The criteria for acceptable
    
                                       C-37
    

    -------
                     TABLE 11.  CHARACTERISTIC SIM IONS FOR PCBs
    
    Homolog
    Ci2H9Cl
    Ci2H8Cl2
    C12H7C13
    C12H6C14
    Ci2H5Cl5
    C12H4C16
    C^HgCly
    C12H2C18
    Ci2HClg
    Ci2Clio
    
    Primary
    188 (100)
    222 (100)
    256 (100)
    292 (100)
    326 (100)
    360 (100)
    394 (100)
    430 (100)
    464 (100)
    498 (100)
    Ion (relative intensity)
    Secondary
    190 (33)
    224 (66)
    258 (99)
    290 (76)
    328 (66)
    362 (82)
    396 (98)
    432 (66)
    466 (76)
    500 (87)
    
    Tertiary
    -
    226 (11)
    260 (33)
    294 (49)
    324 (61)
    364 (36)
    398 (54)
    428 (87)
    462 (76)
    496 (68)
    
    Source:   Rote, J. W.,  and W. J.  Morris, "Use of Isotopic Abundance Ratios in
             Identification of Polychlorinated Biphenyls by Mass Spectrometry,"
             J. Assoc. Offic. Anal.  Chem.,  56(1), 188-199 (1973).
                                         C-38
    

    -------
                TABLE 12.   LIMITED MASS SCANNING (IMS) RANGES FOR PCBs
    
    Compound
    C^Cl,
    C12H8C12
    C12H7C13
    C12H6C13
    C^H5C15
    Ci2H4Cl6
    C12H3C17
    C12H2Clg
    Ci2HCl9
    C12Cl!0
    Ci2D6Cl4
    13C612C6H9C1
    13C12H6C14
    13C12H2C18
    13C12C116
    Mass range (m/z)
    186-190
    220-226
    254-260
    288-294
    322-328
    356-364
    386-400
    426-434
    460-468
    494-504
    294-300
    192-196
    300-306
    438-446
    506-516
    
    a  Adapted from Tindall, G. W.,  and P. E. Wininger, "Gas Chromatography-Mass
       Spectrometry Method for Identifying and Determining Polychlorinated Bi-
       phenyls," J. Chromatogr., 196, 109-119 (1980).
                                          C-39
    

    -------
                         identification include a high index of similarity.   For
                         the Incos 2300, a fit of 750 or greater must be obtained.
    
          11.3  Disputes in interpretation - Where there is  reasonable doubt as
                to the identity of a peak as a PCB,  the analyst must either  iden-
                tify the peak as a PCB or proceed to a confirmational analysis
                (see Section 13.0).
    12.0  Quantitative Data Reduction
    
          12.1  Once a chromatographic peak has  been identified as  a  PCB,  the com-
                pound is quantitated based either on the integrated abundance of
                the SIM data or EICP for the primary characteristic ion in Tables
                11 and 12.   If interferences are observed for the primary  ion,  use
                the secondary and then tertiary  ion for quantitation.   If  inter-
                ferences in the parent cluster prevent quantitation,  an ion from a
                fragment cluster (e.g.,  M-70) may be used.   Whichever ion  is used,
                the RF must be determined using  that ion.   The same criteria
                should be applied to the internal standard compounds  (Table 13).
    
          12.2  Using the appropriate response factor (RF ) as determined  in Sec-
                tion 7.3, calculate the mass of  each PCB peak (M )  using Equation
                12-1.                                           p
    
                                  A      -
                             M  = _E  .   i  . M
                              p   A.    RF    is                       Eq. 12-1
                              *    is     p                             ^
                where    A  = area of the characteristic ion for the  analyte PCB
                          ^     peak
    
                        A.   = area of the characteristic ion for the  internal
                                standard peak
    
                        RF  = response factor of a given PCB congener
    
                        M.   = mass of internal standard injected (micrograms)
                         1S
    
          12.3  If a peak appears to contain non-PCB interferences  which cannot
                be circumvented by a secondary or tertiary ion, either:
    
                12.3.1   Reanalyze the sample on a different column which  sepa-
                         rates the PCB and interferents;
    
                12.3.2   Perform additional chemical cleanup (Section 9) and then
                         reanalyze the sample; or
    
                12.3.3   Quantitate the entire peak as PCB.
    
          12.4  Sum all of the peaks for each homolog and then sum  those to yield
                the total PCB mass, M™,  in the sample.   If a concentration-per-
                peak or concentration-per-homolog reporting format  is desired,
                carry each value through the calculations in an appropriate manner.
    
                                       C-40
    

    -------
            TABLE 13.  CHARACTERISTIC IONS FOR 13C-LABELED PCB SURROGATES
    
    
    
    
                                  	Ion (relative intensity)
    Specific compound              Primary           Secondary          Tertiary
    
    
    
    
    
    
    13C612C6H9C1                  194 (100)          196 (33)
    
    
    
    
    13C12H6C14                    304 (100)          306 (49)           302 (78)
    
    
    
    
    13C12H2C18                    442 (100)          444 (65)           440 (89)
    
    
    
    
    13C12C110                     510 (100)          512 (87)           514 (50)
                                          C-41
    

    -------
    12.5  Calculation of air sample volume1
    
          12.5.1   Nomenclature
                   M  = Mass of PCB represented by a chromatographic peak
                    "     micrograms
    
                      = Total mass of PCBs in sample,  micrograms
    
                   C  = Concentration of PCBs in air,  micrograms per cubic
                          meter,  corrected to standard conditions of 20°C,
                          760 mm Hg (68°F, 29.92 in. Hg)  on dry basis
    
                   A  = Cross-sectional area of nozzle,  square meter (square
                    n     feet)
    
                   B   = Water vapor in the gas stream,  proportion by volume
    
                   I = Percent of isokinetic sampling
    
                   MW  = Molecular weight of water,  18 g/g-mole (18 lb/
                           Ib-mole)
    
                   P,    = Barometric pressure at the sampling site, mm Hg
                    oar     / .    YT \
                            (in.  Hg)
    
                   P  = Absolute stack gas pressure, mm Hg (in. Hg)
                    s
                   P  , = Standard absolute pressure,  760 mm Hg (29.92 in
                    SL.U     TT \
                            Hg)
    
                   R = Ideal gas constant, 0.06236 mm Hg-m3/K-g-mole (21.83 in.
                         Hg-ft3/°R-lb-mole)
    
                   T  = Absolute average dry gas meter temperature °K (°R)
    
                   T  = Absolute average stack gas temperature °K (°R)
                    s
    
                        = Standard absolute temperature,  293°K (528°R)
                   V,   = Total volume of liquid collected in impingers and
                           silica gel, milliliters.   Volume of water col-
                           lected equals the weight  increase in grams times
                           1 ml/g
    
                   V  = Volume of gas sample as measured by dry gas meter,
                    ra     dcm (dcf)
    
                   V ,  ,..  = Volume of gas sample measured by the dry gas
                               meter corrected to standard conditions,
                               dscm (dscf)
                                 C-42
    

    -------
                    V ,   ,x  = Volume  of  water  vapor  in  the  gas  sample  cor-
                      ^          rected to  standard conditions,  scm  (scf)
    
                    V = Total volume of sample, railliliter
    
                    V = Stack gas  velocity, calculated by  EPA  Method  2,
                           m/sec (ft/sec)
    
                    AH = Average pressure  differential  across the orifice
                           meter, mm  H20 (in.
                    p  = Density of water,  1  g/ml  (0.00220  Ib/ml)
                     W
    
                    9 = Total sampling time,  minutes
    
                    13.6 = Specific gravity of  mercury
    
                    60 = Seconds per minute
    
                    100 = Conversion to percent
    
           12.5.2   Average dry gas meter temperature and average  orifice
                    pressure drop - See data  sheet (Figure  4).
    
           12.5.3   Dry gas volume - Correct  the sample  volume  measured by
                    the dry gas meter to standard  conditions  [20°C,  760 mm Hg
                    (68°F, 29.92 in. Hg)} by  using Equation 12-2.
    
    
                            P,   +™          P,
    w           v            bar   13.6     „ 17  *bar   13.6       Eq.  12-2
     m(std) "    m  ~T~       P~T" K  ra     T
      ^   J           m         std                  m
    
           where K = 0.3855°K/mm Hg for metric units
                   = 17.65 °R/in. Hg for English units
    
           12.5.4   Volume of water vapor
    
    
                          = vic ar  ?rf = KVic                 E*-  12'3
                                  w   std
    
           where K = 0.00134 m3/ml for metric units
                   = 0.0472 ft3/ml for English units
    
           12.5.5   Moisture content
    
                  R   -       w(std)                                      ,
                  Bws - V , _, + V f ¥..                         Eq"  12"4
                         m(std)    w(std)
    
                    If the liquid droplets are present  in the  gas  stream,  as-
                    sume the stream to be saturated and use a  psychrometric
                    chart to obtain an approximation  of the moisture per-
                    centage.
                                  C-43
    

    -------
          12.6  Concentration of PCBs in stack gas - Determine the concentration
                of PCBs in the air according to Equation 12-5 and report in micro-
                grams per cubic meter using Table 14.  If an alternate reporting
                format (e.g., concentration per peak) is desired, a different
                report form may be used.
                                 M
                       C  = K  ^ — - -                                  Eq. 12-5
                        a       m(std)
    
                where K = 35.31 ft3/m3
    
          12.7  Isokinetic variation
    
                12.7.1   Calculations from raw data.
    
                 T   100 Ts [K Vlc + (VTm) (Pbar) + M/13'6^         _   „,
                 i = - 6o e v  p  A - —         Eq- 12'6
                                         s  s  n
                where K = 0.00346 mm Hg-m3/ml-°K for metric units
                        = 0.00267 in. Hg-ft3/ml-°R for English units
    
                12.7.2   Calculations from intermediate values
                              T  V , „ ,, P .  , 100
                         -     s  m(std)  std                           _    _ _
                                                                              ~
                           T ^ . V  0 A  P  60 (1-B  )
                            std  s    n  s        ws
                                 T  V
                           v      s  m(std)
                         — K. ---- ' -------
                             P  V  A  0 (1-B  )
                              s  s  n       ws
    
                where K = 4.323 for metric units
                        = 0.0944 for English units
    
                12.7.3   Acceptable results - The following range sets the limit
                         on acceptable isokinetic sampling results:
    
                         If 90% < I < 110%, the results are acceptable.  If the
                         results are low in comparison to the standards and I is
                         beyond the acceptable range,  the Agency may opt to ac-
                         cept the results.
    
          12.8  Round off all numbers reported to two  significant figures.
    
    
    13.0  Confirmation
    
          If there is reason to question the qualitative identification (Section
          11.0), the analyst may choose to confirm that a peak is not a PCB.  Any
          technique may be chosen provided that it is  validated as having equiva-
          lent or superior selectivity and sensitivity to GC/EIMS.  Some candidate
          techniques include alternate GC columns (with EIMS detection), GC/CIMS,
          GC/NCIMS, high resolution EIMS, and MS/MS techniques.  Each laboratory
    
    
                                       C-44
    

    -------
    TABLE 14.  ANALYSIS REPORT
    
    Sample No.
    Sample Matrix
    Sample Source
    INCIDENTAL PCBs IN AIR
    
    
    
    Notebook No. or File Location
    Volume Collected [V , ,-J
    Mass of Internal Standard
    Analyte 1° 2° Il°
    IS 298 246
    1-C1 188 190
    2-C1 222 224
    3-C1 256 258
    4-C1 292 290
    5-C1 326 328
    6-C1 360 362
    7-C1 394 396
    8-C1 430 432
    9-C1 464 466
    10-C1 498 500
    Total (M^
    Concentration (C.)
    A
    Reported by:
    Name
    Signature/Date
    Organization
    m3
    Injected, M._
    Qualitative
    I2° Ratio Theoretical
    100/76
    100/33
    100/66
    100/99
    100/76
    100/66
    100/82
    100/98
    100/66
    100/76
    100/87
    Internal Audit:
    Name
    Signature/Date
    Organization
    M8
    Quantitative
    Ion Mass
    OK? Used RF M (pg)
    P
    1.000
    M8
    jjg/m3
    EPA Audit:
    Name
    Signature/Date
    Organization
                C-45
    

    -------
          must validate confirmation techniques  to show equivalent or superior
          selectivity between PCBs  and interferences  and sensitivity (limit of
          quantitation, LOQ).
    
          If a peak is confirmed as being a non-PCB,  it may be deleted from the
          calculation (Section 12).  If a peak is  confirmed as containing both
          PCB and non-PCB components, it must be quantitated according to Section
          12.3.
    14.0  Quality Control
    
          14.1  Each laboratory that uses this method must operate a formal qual-
                ity control (QC) program.  The minimum requirements of this pro-
                gram consist of an initial demonstration of laboratory capability
                and the analysis of spiked samples  as a continuing check on per-
                formance.  The laboratory must maintain performance records to
                define the quality of data that are generated.   After a date spe-
                cified by the Agency, ongoing performance checks should be  com-
                pared with established performance  criteria to  determine if the
                results of analyses are within accuracy and precision limits ex-
                pected of the method.
    
          14.2  The analysts must certify that the  precision and accuracy of the
                analytical results are acceptable by:
    
                14.2.1   The absolute precision of  surrogate recovery, measured
                         as the RSD  of the integrated EIMS area (A ) for a set
                         of samples, must be ±10%.
    
                14.2.2   The mean recovery (R ) of  at least four replicates of a
                         QC check sample to be supplied by the  Agency must  meet
                         Agency-specified accuracy  and precision criteria.   This
                         forms the initial data base for establishing control
                         limits (see Section 14.3 below).
    
          14.3  Control limits - The laboratory must establish  control limits using
                the following equations:
    
                         Upper control limit (UCL)  = R  + 3 RSD
    
                         Upper warning limit (UWL)  = R  + 2 RSD
    
                         Lower warning limit (LWL)  = R  - 2 RSD
    
                         Lower control limit (LCL)  = R  - 3 RSD
                                                      c        c
                These may be plotted on control charts.  If an  analysis of a check
                sample falls outside the warning limits, the analyst should be
                alerted that potential problems may need correction.  If the re-
                sults for a check sample fall outside the control limits, the lab-
                oratory must take corrective action and recertify the performance
    
    
                                       C-46
    

    -------
          (Section 14.2)  before  proceeding with analyses.   The  warning and
          control limits  should  be continuously updated  as  more check sample
          replicates are  added to the data base.
    
    14.4  Before processing any  samples,  the analyst should demonstrate
          through the analysis of a reagent blank that all  glassware  and
          reagent interferences  are under control.   Each time a set of sam-
          ples is analyzed or there is a  change in reagents,  a  laboratory
          reagent blank should be processed as a safeguard  against contami-
          nation.
    
    14.5  Procedural QC - The various steps of the analytical procedure
          should have quality control measures.  These include  but are not
          limited to:
    
          14.5.1   GC performance - See Section 7.1 for  performance criteria.
    
          14.5.2   MS performance - See Section 7.2 for  performance criteria.
    
          14.5.3   Qualitative identification - At least 10% of the PCB
                   identifications, as well as any questionable results,
                   should be confirmed by a second mass  spectrometrist.
    
          14.5.4   Quantitation  - At least 10% of all manual calculations,
                   including peak area calculation, must be checked.   After
                   changes in computer quantitation routes, the results
                   should be manually checked.
    
    14.6  A minimum of 10% of all samples, one sample per month or one sam-
          ple per matrix type, whichever is greater, must be selected at
          random, sampled, and analyzed in triplicate to monitor the  preci-
          sion of the analysis.   An RSD of ±30% or less  must be achieved.
          If the precision is greater than ±30%, the analyst must be  re-
          certified (see Section 14.2).
    
    14.7  A minimum of 10% of all samples, one sample per month or one sam-
          ple per matrix type, whichever is greater, selected at random,
          must be analyzed by the standard addition technique.   Two aliquots
          of the sample are analyzed, one "as is" and one spiked with a suf-
          ficient amount of solution CSxxx to yield approximately 100 |jg/
          sample of each compound.  The spiking compounds are thoroughly
          incorporated by mechanical agitation.  For the liquid impinger
          contents, shaking for  30 sec should be sufficient.  For the
          Florisil, 10 min tumbling is recommended.  For filters where in-
          adequate incorporation may be expected, overnight equilibration
          with agitation is recommended.
    
          Note:  The volume measurement of the spiking solution is critical
          to the overall method  precision.  The analyst  must exercise cau-
          tion that the volume is known to ±1% or better.  Where necessary,
          calibration of the pipet is recommended.
                                 C-47
    

    -------
                The samples are analyzed together and the quantitative results
                calculated.  The recovery of the spiked compounds (calculated by
                difference) must be 80-120%.   If the sample is known to contain
                specific PCB isomers,  these isomers may be substituted for solu-
                tion CSxxx.  If the concentrations of PCBs are known to be high,
                the amount added should be adjusted so that the spiking level is
                1.5 to 4 times the measured PCB level in the unspiked sample.
    
          14.8  Sampling efficiency -  The efficiency of PCB collection during
                sampling should be monitored.   This may be achieved by adding a
                known amount of the 13C surrogate spiking solution (Section 6.4)
                sufficient to give an  analytical signal well above background to
                the first impinger prior to sampling.  The recovery of the four
                compounds should be >  80%.
    
          14.9  Interlaboratory comparison - Interlaboratory comparison studies
                are planned.  Participation requirements, level of performance,
                and the identity of the coordinating laboratory will be presented
                in later revisions.
    
          14.10 It is recommended that the participating laboratory adopt addi-
                tional QC practices for use with this method.   The specific prac-
                tices that are most productive depend upon the needs of the lab-
                oratory and the nature of the samples.  Field duplicates or
                triplicates may be analyzed to monitor the precision of the sam-
                pling technique.  Whenever possible, the laboratory should per-
                form analysis of standard reference materials  and participate in
                relevant performance evaluation studies.
    
    
    15.0  Quality Assurance
    
          Each participating laboratory must develop a quality assurance plan ac-
          cording to EPA guidelines.5   The quality assurance plan must be submitted
          to the Agency for approval.
    
    
    16.0  Method Performance
    
          The method performance is being evaluated.  Limits of quantitation;
          average intralaboratory recoveries,  precision, and accuracy; and inter-
          laboratory recoveries, precision, and accuracy will be presented.
    
    
    17.0  Documentation and Records
    
          Each laboratory is responsible for maintaining full records of the analy-
          sis.  Laboratory notebooks should be used for handwritten records. GC/MS
          data must be archived on magnetic tape, disk, or a similar device. Hard
          copy printouts may be kept in addition if desired.  QC records should
          be maintained separately from sample analysis records.
                                       C-48
    

    -------
    The documentation must describe completely how the analysis was performed.
    Any variances from the protocol must be noted and fully described.  Where
    the protocol lists options (e.g., sample cleanup), the option used and
    specifies (solvent volumes, digestion times, etc.) must be stated.
                                 C-49
    

    -------
                                     REFERENCES
    
    1.  Haile, C. L.,  and E. Baladi, "Methods for Determining the Polychlorinated
        Biphenyl Emissions from Incineration and Capacitor and Transformer Filling
        Plants," U.S.  Environmental Protection Agency,  (1977) EPA-600/4-73-048.
    
    2.  U.S. Environmental Protection Agency, Federal Register, 4j2(160),  Thursday,
        August 18, 1977.
    
    3.  Martin, R. M., "Construction Details of Isokinetic Source Sampling Equip-
        ment," Environmental Protection Agency, Air Pollution Control Office
        Publication No. APTD-0581.
    
    4.  Bellar, T. A., and J. J. Lichtenberg, "The Determination of Polychlorinated
        Biphenyls in Transformer Fluid and Waste Oils," Prepared for U.S.  Environ-
        mental Protection Agency, (1981) EPA-600/4-81-045.
    
    5.  Quality Assurance Program Plan for the Office of Toxic Substances, Office
        of Pesticides  and Toxic Substances, U.S. Environmental Protection Agency,
        Washington, D.C., October 1980.
                                       C-50
    

    -------
                            APPENDIX D
    
    
    ANALYTICAL METHOD:  THE ANALYSIS OF BY-PRODUCT CHLORINATED
                BIPHENYLS IN INDUSTRIAL WASTEWATER
                              D-l
    

    -------
                  THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS
                              IN INDUSTRIAL WASTEWATER
    1.0   Scope and Application
    
          1.1    This is  a  gas  chromatographic/electron  impact  mass  spectrometric
                (GC/EIMS)  method  applicable  to  the  determination  of chlorinated
                biphenyls  (PCBs)  in  industrial  wastewater.   The PCBs present  may
                originate  either  as  synthetic by-products or as contaminants  de-
                rived from commercial  PCB  products  (e.g., Aroclors).   The  PCBs
                may be present as single isomers  or complex  mixtures and may  in-
                clude all  209  congeners from monochlorobiphenyl through deca-
                chlorobiphenyl listed  in Table  1.
    
          1.2    The detection  and quantitation  limits are dependent upon the  vol-
                ume of sample  extracted the  complexity  of the  sample matrix and
                the ability of the analyst to remove interferents and  properly
                maintain the analytical system.   The method  accuracy and preci-
                sion will  be determined in future studies.
    
          1.3    This method is restricted  to use  by or  under the  supervision  of
                analysts experienced in the  use of  gas  chromatography/mass spec-
                trometry (GC/MS)  and in the  interpretation of  gas chromatograms
                and mass spectra.  Prior to  sample  analysis, each analyst  must
                demonstrate the ability to generate acceptable results with this
                method by  following  the procedures  described in Section 14.2.
    
          1.4    The validity of the  results  depends on  equivalent recovery of the
                analyte  and 13C PCBs.  If  the *3C PCBs  are not thoroughly  incor-
                porated  in the matrix, the method is not applicable.
    
          1.5    During the development and testing  of this method,  certain analyti-
                cal parameters and equipment designs were found to  affect  the valid-
                ity of the analytical  results.  Proper  use of  the method requires
                that such  parameters or designs must be used as specified.  These
                items are  identified in the  text  by the word "must."  Anyone  wish-
                ing to deviate from  the method  in areas so identified  must demon-
                strate that the deviation  does  not  affect the  validity of  the data.
                Alternative test  procedure approval must be  obtained from  the
                Agency.  An experienced analyst may make modifications to  param-
                eters or equipment identified by  the term "recommended."   Each
                time such  modifications are  made  to the method, the analyst must
                repeat the procedure in Section 14.2.   In this case, formal ap-
                proval is  not  required, but  the documented data from Section  14.2
                must be  on file as part of the  overall  quality assurance program.
                                       D-2
    

    -------
                                   TABLE 1.  NUMBERING OF  PCB CONGENERS3
    NO.
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    IS
    IS
    17
    13
    19
    20
    21
    22
    23
    24
    25
    26
    27
    28
    29
    30
    31
    32
    33
    34
    35
    36
    37
    38
    39
    40
    41
    42
    43
    44
    45
    4«
    47
    48
    49
    50
    51
    Structure
    Mo«aeMorob
    -------
    2.0   Summary
          2.1   The wastewater must be sampled such that the specimen collected
                for analysis is representative of the whole.  Statistically
                designed selection of the sampling position (valve, port, outfall,
                etc.) or time should be employed.  The sample must be preserved to
                prevent PCB loss prior to analysis.  Storage at 4°C with optional
                preservation at low pH is recommended.
    
          2.2   The sample is mechanically homogenized and subsampled if necessary.
                The sample is then spiked with four 13C PCB surrogates and the
                surrogates incorporated by further mechanical agitation.
    
          2.3   The surrogate-spiked sample is extracted and cleaned up at the
                discretion of the analyst.  Possible extraction techniques include
                liquid-liquid partition and sorption onto resin columns followed
                by solvent elution.   Cleanup techniques may include liquid-liquid
                partition, sulfuric acid cleanup, saponification,  adsorption chro-
                matography, gel permeation chromatography or a combination of
                cleanup techniques.   The sample is diluted or concentrated to a
                final known volume for instrumental determination.   The EPA Method
                6081 and 6252 extraction and cleanup procedures may be used.
    
          2.4   The PCB content of the sample extract is determined by capillary
                (preferred) or packed column gas chromatography/electron impact
                mass spectrometry (CGC/EIMS or PGC/EIMS) operated  in the selected
                ion monitoring (SIM), full scan, or limited mass scan (LMS) mode.
    
          2.5   PCBs are identified by comparison of their retention time and
                mass spectral intensity ratios to those in calibration standards.
    
          2.6   PCBs are quantitated against the response factors  for a mixture
                of 11 PCB congeners, using the response of the 13C surrogate to
                compensate for losses in workup and instrument variability.
    
          2.7   The PCBs identified by the SIM technique may be confirmed by full
                scan CGC/EIMS, retention on alternate GC columns,  other mass spec-
                trometric techniques, infrared spectrometry, or other techniques,
                provided that the sensitivity and selectivity of the technique is
                demonstrated to be comparable or superior to GC/EIMS.
    
          2.8   The analysis time is dependent on the extent of workup employed.
                The time required for instrumental analysis, excluding data re-
                duction and reporting, is about 30 to 45 min.
    
          2.9   Appropriate quality control (QC) procedures are included to assess
                the performance of the analyst and estimate the quality of the
                results.  These QC procedures include the demonstration of labora-
                tory capability:   periodic analyst certification,  the use of con-
                trol charts, and the analysis of blanks, replicates, and standard
                addition samples.  A quality assurance (QA) plan must be developed
                for each laboratory.
                                       D-4
    

    -------
          2.10  While several options are available throughout this method, the
                recommended procedure to be followed is:
    
                2.10.1   The sample is collected according to a scheme which per-
                         mits extrapolation of the sample data to the body or con-
                         tainers of water being sampled.
    
                2.10.2   The sample is preserved at low pH and at 4°C to prevent
                         any loss of PCBs or changes in matrix which may adversely
                         affect recovery.
    
                2.10.3   The sample is-mechanically homogenized and subsampled if
                         necessary.
    
                2.10.4   The sample is spiked with four 13C-PCB surrogates
                         (4-chlorobiphenyl; 3,3',4,4'-tetrachlorobiphenyl;
                         2,2',3,3',5,5',6,6'-octachlorobiphenyl; and decachloro-
                         biphenyl).
    
                2.10.5   The sample is extracted.
    
                2.10.6   The extract is cleaned up and concentrated to an appro-
                         priate volume.
    
                2.10.7   An aliquot of the extract is  analyzed by CGC/EIMS oper-
                         ated in the SIM mode.  On-column injections onto a 15-m
                         DB-5 capillary column, programmed (for toluene solutions)
                         from 110°  to 325°C at 10°/min after a 2 rain hold is used.
                         Helium at  45-cm/sec linear velocity is used as the carrier
                         gas.
    
                2.10.8   PCBs are identified by retention time and mass spectral
                         intensities.
    
                2.10.9   PCBs are quantitated against  the response factors for a
                         mixture of 11 PCB congeners.
    
                2.10.10  The total  PCBs are obtained by summing the amounts for
                         each homolog found and the concentration is reported as
                         micrograms per liter.
    
    
    3.0   Interferences
    
          3.1   Method interferences may be caused by  contaminants in solvents,
                reagents, glassware, and other sample  processing hardware, leading
                to discrete artifacts and/or elevated  baselines in the total ion
                current profiles.  All of these materials must be routinely demon-
                strated to be free  from interferences  by the analysis of laboratory
                reagent blanks as described in Section 14.4.
                                       D-5
    

    -------
    4.0
                3.1.1    Glassware must be scrupulously cleaned.   All glassware
                         is cleaned as soon as possible after use by rinsing with
                         the last solvent used.  This should be followed by deter-
                         gent washing with hot water and rinses with tap water and
                         reagent water.  The glassware should then be drained dry
                         and heated in a muffle furnace at 400°C  for 15 to 30 min.
                         Some thermally stable materials,  such as PCBs, may not
                         be eliminated by this treatment.   Solvent rinses with
                         acetone and pesticide quality hexane may be substituted
                         for the muffle furnace heating.   Volumetric ware should
                         not be heated in a muffle furnace.   After it is dry and
                         cool, glassware should be sealed  and stored in a clean
                         environment to prevent any accumulation  of dust or other
                         contaminants.  It is stored inverted or  capped with
                         aluminum foil.
    
                3.1.2    The use of high purity reagents  and solvents helps to
                         minimize interference problems.   Purification of solvents
                         by distillation in all-glass systems may be required.
                         All solvent lots must be checked  for purity prior to use.
    
          3.2   Matrix interferences may be caused by contaminants that are coex-
                tracted from the sample.  The extent of matrix interferences will
                vary considerably from source to source,  depending upon the nature
                and diversity of the sources of samples.
          4.1   The toxicity or carcinogenicity of each reagent used in this
                method has not been precisely defined;  however, each chemical
                compound should be treated as a potential health hazard.   From
                this viewpoint, exposure to these chemicals must be reduced to
                the lowest possible level by whatever means available.   The labor-
                atory is responsible for maintaining a  current awareness  file of
                OSHA regulations regarding the safe handling of the chemicals spe-
                cified in this method.   A reference file of material data handling
                sheets should also be made available to all personnel involved in
                the chemical analysis.
    
          4.2   Polychlorinated biphenyls have been tentatively classified as known
                or suspected human or mammalian carcinogens.   Primary standards
                of these toxic compounds should be prepared in a hood.   Personnel
                must wear protective equipment, including gloves and safety glasses.
    
                Congeners highly substituted at the meta and para positions and
                unsubstituted at the ortho positions are reported to be the most
                toxic.  Extreme caution should be taken when handling these com-
                pounds neat or in concentration solution.   The class includes
                3,3',4,4'-tetrachlorobiphenyl (both natural abundance and isotop-
                ically labeled).
                                       D-6
    

    -------
          4.3   Diethyl ether should be monitored regularly to determine the perox-
                ide content.   Under no circumstances should diethyl ether be used
                with a peroxide content in excess of 50 ppm as an explosion could
                result.  Peroxide test strips manufactured by EM Laboratories
                (available from Scientific Products Company,  Cat. No.  P1126-8 and
                other suppliers) are recommended for this test.   Procedures for
                removal of peroxides from diethyl ether are included in the in-
                structions supplied with the peroxide test kit.
    
          4.4   Waste disposal must be in accordance with RCRA and applicable
                state rules.
    
    5.0   Apparatus and Materials
    
          5.1   Sampling containers - Amber glass bottles, 1-liter or other ap-
                propriate volume, fitted with screw caps lined with Teflon.
                Cleaned foil  may be substituted for Teflon if the sample is not
                corrosive.  If amber bottles are not available,  samples should
                be protected  from light using foil or a light-tight outer con-
                tainer.  The  bottle must be washed, rinsed with acetone or methy-
                lene chloride, and dried before use to minimize contamination.
    
          5.2   Glassware - All specifications are suggestions only.  Catalog
                numbers are included for illustration only.
    
                5.2.1    Volumetric flasks - Assorted sizes.
    
                5.2.2    Pipets - Assorted sizes, Mohr delivery.
    
                5.2.3    Micro syringes - 10.0 pi for packed  column GC analysis,
                         1.0  pi for on-column CGC analysis.
    
                5.2.4    Chromatographic column - Chromaflex, 400 mm long x 19 mm
                         ID (Kontes K-420540-9011 or equivalent).
    
                5.2.5    Gel  permeation chromatograph - GPC Autoprep 1002
                         (Analytical Bio Chemistry Laboratories,  Inc.) or
                         equivalent.
    
                5.2.6    Kuderna-Danish Evaporative Concentrator Apparatus
    
                         5.2.6.1  Concentrator tube - 10 ml,  graduated (Kontes
                                  K-570050-1025 or equivalent).   Calibration must
                                  be checked.  Ground glass stopper size (S19/22
                                  joint) is used to prevent evaporation of solvent.
    
                         5.2.6.2  Evaporative flask - 500 ml  (Kontes K-57001-0500
                                  or equivalent).  Attach to  concentrator tube
                                  with springs (Kontes K-662750-0012 or equivalent),
    
                         5.2.6.3  Snyder column - Three ball  macro (Kontes K503000-
                                  0121 or equivalent).
    
    
                                       D-7
    

    -------
    5,3   Balance - Analytical, capable of accurately weighing 0.0001 g.
    
    5.4   Gas chromatography/mass spectrometer system.
    
          5.4.1    Gas chromatograph - An analytical system complete with a
                   temperature programmable gas chromatograph and all re-
                   quired accessories including syringes,  analytical columns,
                   and gases.  The injection port must be  designed for on-
                   column injection when using capillary columns or packed
                   columns.   Other capillary injection techniques (split,
                   splitless, "Grob," etc.) may be used provided the per-
                   formance  specifications stated in Section 7.1 are met.
    
          5.4.2    Capillary GC column - A 12-20 m long x  0.25 mm ID fused
                   silica column with a 0.25 |Jm thick DB-5 bonded silicone
                   liquid phase (J&W Scientific) is recommended.  Alternate
                   liquid phases may include OV-101, SP-2100,  Apiezon L,
                   Dexsil 300, or other liquid phases which meet the perfor-
                   mance specifications stated in Section  7.1.
    
          5.4.3    Packed GC column - A 180 cm x 0.2 cm ID glass column
                   packed with 3% SP-2250 on 100/120 mesh  Supelcoport or
                   equivalent is recommended.   Other liquid phases which
                   meet the  performance specifications stated in Section  7.1
                   may be substituted.
    
          5.4.4    Mass spectrometer - Must be capable of  scanning from 150
                   to 550 Daltons every 1.5 sec or less, collecting at least
                   five spectra per chromatographic peak,  utilizing a 70-eV
                   (nominal) electron energy in the electron impact ioniza-
                   tion mode and producing a mass spectrum which meets all
                   the criteria in Table 2 when 50 ng of decafluorotriphenyl
                   phosphine [DFTPP, bis(perfluorophenyl)phenyl phosphine]
                   is injected through the GC inlet.  Any  GC-to-MS interface
                   that gives acceptable calibration points at 10 ng per
                   injection for each PCB isomer in the calibration standard
                   and achieves all acceptable performance criteria (Section
                   10) may be used.  Direct coupling of the fused silica
                   column to the MS is recommended.  Alternatively, GC-to-
                   MS interfaces constructed of all glass  or glass-lined
                   materials are recommended.   Glass can be deactivated by
                   silanizing with dichlorodimethylsilane.
    
          5.4.5    A computer system that allows the continuous acquisition
                   and storage on machine-readable media of all mass spectra
                   obtained  throughout the duration of the chromatographic
                   program must be interfaced to the mass  spectrometer.
                   The data  system must have the capability of integrating
                   the abundances of the selected ions between specified
                   limits and relating integrated abundances to concentra-
                   tions using the calibration procedures  described in this
                   method.   The computer must have software that allows
                                 D-8
    

    -------
       TABLE 2.  DFTPP KEY IONS AND ION ABUNDANCE CRITERIA	
    
    Mass                 .          Ion abundance criteria
    
    
    197                            Less than 1% of mass 198
    198                            100% relative abundance
    199                            5-9% of mass 198
    
    275                            10-30% of mass 198
    
    365                            Greater than 1% of mass 198
    
    441                            Present, but less than mass 443
    442                            Greater than 40% of mass 198
    443                            17-23% of mass 442
                               D-9
    

    -------
                         searching any GO/MS data file for ions of a specific mass
                         and plotting such ion abundances versus time or scan num-
                         ber to yield an extracted ion current profile (EICP).
                         Software must also be available that allows integrating
                         the abundance in any EICP between specified time or scan
                         number limits.
    6.0   Reagents
          6.1   Solvents - All solvents must be pesticide residue analysis grade.
                New lots should be checked for purity by concentrating an aliquot
                by at least as much as is used in the procedure.
    
          6.2   Stock standard solutions - Standards of the PCB congeners listed
                in Table 3 are available from Ultra Scientific, Hope, Rhode Island;
                or Analabs, North Haven, Connecticut.
    
          6.3   Calibration standard stock solutions - Primary dilutions of each
                of the individual PCBs listed in Table 3 are prepared by weighing
                approximately 1-10 mg of material within 1% precision.  The PCB
                is then dissolved and diluted to 1.0 ml with hexane.   Calculate
                the concentration in mg/ml.   The primary dilutions are stored at
                4°C in screw-cap vials with Teflon cap liners.  The meniscus is
                marked on the vial wall to monitor solvent evaporation.   Primary
                dilutions are stable indefinitely if the seals are maintained.
                The validity of primary and secondary dilutions must be monitored
                on a quarterly basis by analyzing four quality control check sam-
                ples (see Section 14.2).
    
          6.4   Working calibration standards - Working calibration standards are
                prepared that are similar in PCB composition and concentration to
                the samples by mixing and diluting the individual standard stock
                solutions.  Example calibration solutions are shown in Table 3.
                The mixture is diluted to volume with pesticide residue analysis
                quality hexane.  The concentration is calculated in ng/ml as the
                individual PCBs.  Dilutions are stored at 4°C in narrow-mouth,
                screw-cap vials with Teflon cap liners.  The meniscus is marked
                on the vial wall to monitor solvent evaporation.   These secondary
                dilutions can be stored indefinitely if the seals are maintained.
                These solutions are designated "CSxxx," where the xxx is used to
                encode the nominal concentration in ng/ml.
    
          6.5   Alternatively, certified stock solutions similar to those listed
                in Table 3 may be available from a supplier, in lieu of the pro-
                cedures described in Section 6.4.
    
          6.6   DFTPP standard - A SO-ng/pl solution of DFTPP is prepared in ace-
                tone or another appropriate solvent.
    
          6.7   Surrogate standard stock solution - The four 13C-labeled PCBs
                listed in Table 4 may be available from a supplier as a certified
                solution.  This solution may be used as received or diluted
                further.  These solutions are designated "SSxxx," where the xxx
                is used to encode the nominal concentration in ng/ml.
    
                                       D-10
    

    -------
     TABLE 3.  CONCENTRATIONS OF CONGENERS IN PCS CALIBRATION STANDARDS (ng/ml)a
    
    Homo log
    1
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    4
    1
    4
    8
    10
    Congener
    no.
    1
    3
    7
    30
    50
    :97
    143
    183
    202
    207
    209
    210 (IS)
    211 (RS)
    212 (RS)
    213 (RS)
    214 (RS)
    CS1000
    1,040
    1,000
    1,040
    1,040
    1,520
    1,740
    1,920
    2,600
    4,640
    5,060
    4,240
    255
    104
    257
    407
    502
    CS100
    104
    100
    104
    104
    152
    174
    192
    260
    464
    506
    424
    255
    104
    257
    407
    502
    CS050
    52
    50
    52
    52
    76
    87
    96
    130
    232
    253
    212
    255
    104
    257
    407
    502
    CS010
    10
    10
    10
    10
    15
    17
    19
    26
    46
    51
    42
    255
    104
    257
    407
    502
    
    a  Concentrations given as examples only.
                                     D-ll
    

    -------
            TABLE 4.  COMPOSITION OF SURROGATE SPIKING SOLUTION (SS100)
                             CONTAINING 13C-LABELED PCBsa
    
    Congener
    no.
    211
    212
    213
    214
    Compound
    (I1 ,2' ,3' ,4' ,5' ,6'-13C6)4-chlorobiphenyl
    (13C12)3,3' ,4,4'-tetraqhlorobiphenyl
    (13C12)2,2' ,3,3' ,5,5' ,6,6'-octachlorobiphenyl
    ( 1 3C ! 2 ) decachlorobiphenyl
    Concentration
    (Mg/ml)
    104
    257
    395
    502
    
    a  Concentrations given as examples only.
                                      D-12
    

    -------
          6.8   Internal standard solution - A solution of de-3,3' ,4,4'-tetra-
                chlorobiphenyl is prepared at a nominal concentration of 1-10
                rag/ml in hexane.   The solution is further diluted to give a work-
                ing standard.
    
          6.9   Solution stability - The calibration standard,  surrogate and
                DFTPP solutions should be checked frequently for stability.
                These solutions should be replaced after 6 months,  or sooner if
                comparison with quality control check samples indicates compound
                degradation or concentration change.
    
          6.10  Quality control check samples will be supplied  by the Agency.
    
    
    7.0   Calibration
    
          7.1   The gas chromatograph must meet the minimum operating parameters
                shown in Tables 5 and 6, daily.  If all of the  criteria are not
                met, the analyst must adjust conditions and repeat the test until
                all criteria are met.
    
          7.2   The mass spectrometer must meet the minimum operating parameters
                shown in Tables 2, 7, and 8, daily.  If all criteria are not met,
                the analyst must retune the spectrometer and repeat the test un-
                til all conditions are met.
    
          7.3   The PCB response factor (RF ) must be determined using Equation
                7-1 for the analyte homologi.
    
                                A  x M.
    
                          "p = £^
    
                where    RF  = response factor of a given PCB isomer
    
                          A  = area of the characteristic ion for the PCB congener
                           ^     peak
    
                          M  = mass of PCB congener injected (nanograms)
    
                         A.  = area of the characteristic ion for the internal
                          18     standard peak
    
                         M.  = mass of internal standard injected (nanograms)
                          IS
    
                Using the same conditions as for RF , the surrogate response
                factors (RF ) must be determined using Equation 7-2.
    
                                  A  x M.
                                   IS    S
    
                where A  = area of the characteristic ion for the surrogate peak
                       S
    
                      M  = mass of surrogate injected (nanograms)
                       S
    
                Other items are the same as defined in Equation 7-1.
                                       D-13
    

    -------
    TABLE 5.  OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM
            Parameter
          Recommended
         Tolerance
    Gas chromatograph
    Column
    
    Liquid phase
    
    Liquid phase thickness
    Carrier gas
    Carrier gas velocity
    Injector
    Injector temperature
    Injection volume
    Initial column temperature
    Column temperature program
    Finnigan 9610
    15 m x 0.255 mm ID
    Fused silica
    DB-5 (J&W)
    0.25 \M
    Helium
    /c    ,   b
    45 cm/sec
                   p
    On-column (J&W)
    Optimum performance
    1.0 |jlc
    70°C (2 min)d
    70°-325°C at 10°C/min£
     Other
     Other
    
     Other nonpolar
     or semipolar
     < 1 |Jm
     Hydrogen
     Optimum performance
     Other
     Optimum performance
     Other
     Other
    Other
    Separator
    Transfer line temperature
    Tailing factor
    Peak width
    None
    280°C
    0.7-1.5
    7-10 sec
    Glass jet or othe
    Optimum*
    0.4-3
    < 15 sec
    
    a  Substitutions permitted with any common apparatus or technique provided
       performance criteria are met.
    b  Measured by injection of air or methane at 270°C oven temperature.
    c  For on-column injection, manufacturer's instructions should be followed
       regarding injection technique.
    d  With on-column injection, initial temperature equals boiling point of the
       solvent; in this instance, hexane.
    e  C12C11Q elutes at 270°C.  Programming above this temperature ensures a
       clean column and lower background on subsequent runs.
    f  Fused silica columns may be routed directly into the ion source to pre-
       vent separator discrimination and losses.
    g  High enough to elute all PCBs,  but not high enough to degrade the column
       if routed through the transfer line.
    h  Tailing factor is width of front half of peak at 10% height divided by width
       of back half of peak at 10% height for single PCB congeners in solution CSxxx.
    i  Peak width at 10% height for a single PCB congener is CSxxx.
                                       D-14
    

    -------
     TABLE 6.  OPERATING PARAMETERS FOR PACKED COLUMN GAS CHROMATOGRAPHY SYSTEM
          Parameter
       Recommended
       Tolerances
    Gas chromatograph
    Column
    Finnigan 9610
    180 cm x 0.2 cm ID
    Other3
    Other
    Column packing
    
    
    Carrier gas
    
    Carrier gas flow.rate
    
    Injector
    
    Injector temperature
    
    Injection volume
    
    Initial column temperature
    
    Column temperature program
    
    Separator
    
    Transfer line temperature
    
    Tailing factor
    
    Peak width
    glass
    
    3% SP-2250 on 100/
    120 mesh Supelcoport
    
    Helium
    
    30 ml/min
    
    On-column
    
    250°C
    
    1.0 |jl
    
    150°C, 4 min
    
    150°C-260° at 8°/min
    
    Glass jet
    
    280°C
    
    0.7-1.5
    
    10-20 sec
    Other nonpolar
    or semipolar
    
    Hydrogen
    
    Optimum performance
    
    
    
    Optimum
    
    ^ 5 |Jl
    
    Other
    
    Other
    
    Other
    
    Optimum
    
    0.4-3
    
    < 30 sec
    a  Substitutions permitted if performance criteria are met.
    
    b  High enough to elute all PCBs.
    
    c  Tailing factor is width of front half of peak at 10% height divided by
       width of back half of peak at 10% height for single PCB congeners in solu-
       tion CSxxx.
    
    d  Peak width at 10% height for a single PCB congener in CSxxx.
                                      D-15
    

    -------
       TABLE 7.  OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER SYSTEM
    
          Parameter                    Recommended                Tolerance
    Mass spectrometer
    Data system
    Scan range
    Scan time
    Resolution
    Ion source temperature
    Electron energy
    Trap current
    Multiplier voltage
    Preamplifier sensitivity
    Finnigan 4023
    Incos 2400
    95-550
    1 sec
    Unit
    280°C
    70 eV
    0.2 mA
    -1,600 V
    10"6 A/V
    Other3
    Other
    Other
    Otherb
    Optimum performance
    200°-300°C
    Optimum performance
    Optimum performance
    Optimum performance
    Set for desired
    working range
    
    a  Substitutions permitted if performance criteria are met.
    
    b  Greater than five data points over a GC peak is a minimum.
    
    c  Filaments should be shut off during solvent elution to improve instrument
       stability and prolong filament life, especially if no separator is used.
                                     D-16
    

    -------
     TABLE 8.  OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER SYSTEM
    
    Parameter
    Mass spectrometer
    Data system
    Scan range
    Scan mode
    Cycle time
    Resolution
    Ion source temperature
    Electron energy
    Emission current
    Filament current
    Multiplier
    Recommended
    Finnigan MAT 31 1A
    Incos 2400
    98-550
    Exponential
    1.2 sec
    1,000
    280°C
    70 eV
    1-2 mA
    Optimum
    -1,600 V
    Tolerance
    Other3
    Other
    Other
    Other
    Otherb
    > 500
    250°-300°C
    70 eV
    Optimum
    Optimum
    Optimum
    
    a  Substitutions permitted if performance criteria are met.
    
    b  Greater than five data points over a GC peak is a minimum.
    
    c  Filaments should be shut off "during solvent elution to improve instrument
       stability and prolong filament life, especially if no separator is used.
                                       D-17
    

    -------
                If specific congeners are known to be present and if standards
                are available, selected RF values may be employed.   For general
                samples, solutions CSxxx and SSxxx or a mixture (Tables 3 and 4),
                with a similar level of internal standard (de-3,31,4,4'-tetra-
                chlorobiphenyl) added,  may be used as the response  factor solution.
                The PCB-surrogate pairs to be used in the RF calculation are listed
                in Table 9.
    
                Generally,  only the primary ions of both the analyte and surrogate
                are used to determine the RF values.  If alternate  ions are to be
                used in the quantitation, the RF must be determined using that
                characteristic ion.
    
                The RF value must be determined in a manner to assure ±20% accu-
                racy and precision.  For instruments with good day-to-day preci-
                sion, a running mean (RF) based on seven values determined once
                each day may be appropriate.  Other options include, but are not
                limited to, triplicate determinations of a single concentration
                spaced throughout a day or determination of the RF  at three dif-
                ferent levels to establish a working curve.
    
                If replicate RF values differ by greater than ±10%  RSD, the system
                performance should be monitored closely.  If the RSD is greater
                than ±20%,  the data set must be considered invalid  and the RF re-
                determined  before further analyses are done.
    
          7.4   If the GC/EIMS system has not been demonstrated to  yield a linear
                response or if the analyte concentrations are more  than two orders
                of magnitude different from those in the RF solution, a calibration
                curve must  be prepared.  If the analyte and RF solution concentra-
                tions differ by more than one order of magnitude, a calibration
                curve should be prepared.  A calibration curve should be estab-
                lished with triplicate determinations at three or more concentra-
                tions bracketing the_ analyte levels.
    
          7.5   The relative retention time (RRT) windows for the 10 homologs and
                surrogates  must be determined.  If all congeners are not available,
                a mixture of available congeners or an Aroclor mixture (e.g.,
                1016/1254/1260) may be used to estimate the windows.  The windows
                must be set wider than observed if all isomers are  not determined.
                Typical RRT windows for one column are listed in Table 10.  The
                windows may differ substantially if other GC parameters are used.
    
    
    8.0   Sample Collection, Handling,  and Preservation
    
          8.1   Amber glass sample containers should have Teflon-lined screw caps.
                With noncorrosive samples, methylene chloride-washed aluminum foil
                liners may  be substituted.  The volume is determined by the amount
                of sample to be collected but will usually be 1 liter or 1 qt.
                The sample  size is dependent on the anticipated PCB levels and
                difficulty  of the subsequent extraction/cleanup steps.
    
    
                                       D-18
    

    -------
                             TABLE 9.  PAIRINGS OF ANALYTE, CALIBRATION, AND SURROGATE COMPOUNDS
    vO
    
    Analyte
    Congener
    no.
    1
    2,3
    4-15
    16-39
    40-81
    82-127
    128-169
    170-193
    194-205
    206-208
    209
    Compound
    Calibration standard
    Congener
    no.
    2"Ci2HgCl 1
    3- and 4-C12H9Cl 3
    C12H8C12
    C12H7C13
    C12H6C14
    C12H5C15
    
    C12H3C17
    C12H2C18
    C12HC19
    C12C110
    7
    30
    50
    97
    143
    183
    202
    207
    209
    2
    4
    2
    2
    2
    2
    2
    2
    2
    2
    C
    
    ,4
    ,4,
    ,2'
    ,2'
    ,2'
    ,2'
    ,2'
    ,2'
    12C
    
    
    6
    ,4
    ,3
    ,3
    ,3
    ,3
    ,3
    :ii
    Compound
    
    
    
    ,6
    i
    ,4
    i
    ,3
    ,3
    0
    
    
    
    
    4,5
    ,5,6'
    4, 4', 5', 6
    ',5,5', 6, 6'
    ',4, 4', 5, 6, 6'
    
    Surrogate
    Congener
    no.
    211
    211
    211
    212
    212
    212
    212
    213
    213
    213
    214
    Compound
    13C6-4
    13C6-4
    13C6-4
    13Ci2-3,3'
    13C12-3,3'
    13C12-3,3'
    13C12-3,3'
    13Ci2-2,2'
    13Ci2-2,2'
    13C12-2,2'
    13C12Clio
    
    
    ,4,4'
    ,4,4'
    ,4,4'
    ,4,4'
    ,3,3'
    0 01
    ,J,J
    0 01
    ,0,0
    
    
    
    
    
    
    
    ,5, 5', 6, 6'
    ,5, 5', 6, 6'
    ,5, 5', 6, 6'
    
    
        a  Ballschmiter numbering system, see Table 1.
    

    -------
           TABLE 10.  RELATIVE RETENTION TIME (RRT) RANGES OF PCB HOMOLOGS
                       VERSUS d6-3,3'.4,4'-TETRACHLOROBIPHENYL
    
    PCB
    homolog
    Monochloro
    
    Dichloro
    Trichloro
    Tetrachloro
    Pentachloro
    Hexachloro
    Heptachloro
    Octachloro
    Nonachloro
    Decachloro
    No. of
    isomers
    measured
    3
    
    10
    9
    16
    12
    13
    4
    6
    3
    1
    Calibration solution
    Observed range
    of RRTsa
    0.40-0.50
    
    0.52-0.69
    0.62-0.79
    0.72-1.01
    0.82-1.08
    0.93-1.20
    1.09-1.30
    1.19-1.36
    1.31-1.42
    1.44-1.45
    Congener
    no.
    1
    3
    7
    30
    50
    97
    143
    183
    202
    207
    209
    Observed
    RRT3
    0.43
    0.50
    0.58
    0.65
    0.75
    0.98
    1.05
    1.15
    1.19
    1.33
    1.44
    Projected
    range of
    RRTsD
    0.35-0.55
    
    0.35-0.80
    0.35-1.10
    0.55-1.05
    0.80-1.10
    0.90-1.25
    1.05-1.35
    1.10-1.50
    1.25-1.50
    1.35-1.50
    
    a  The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were
       measured versus 3,3',4,4'-tetrachlorobiphenyl-de (internal standard) using
       a 15-m J&W DB-5 fused silica 'column with a temperature program of 110°C
       for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an on-
       column injector.  A Finnigan 4023 Incos quadrupole mass spectrometer oper-
       ating with a scan range of 95-550 daltons was used to detect each PCB
       congener.
    
    b  The projected relative retention windows account for overlap of eluting
       homologs and take into consideration differences in operating systems and
       lack of all possible 209 PCB congeners.
                                        D-20
    

    -------
          8.2   Sample bottle preparation
    
                8.2.1     All sample  bottles  and  caps  should be  washed  in detergent
                         solution, rinsed with tap water  and  then with distilled
                         water.   The bottles and caps are allowed to drain dry in
                         a  contaminant-free  area.  Then the caps are rinsed with
                         pesticide grade  hexane  and allow to  air dry.
    
                8.2.2     Sample  bottles are  heated to 400°C for 15  to  20 min or
                         rinsed  with pesticide grade  acetone  or hexane and allowed
                         to air  dry.
    
                8.2.3     The clean bottles are stored inverted  or sealed until use.
    
          8.3   Sample collection
    
                8.3.1     The primary consideration in sample  collection is that
                         the sample  collected be representative of  the whole.
                         Therefore,  sampling plans or protocols for each individ-
                         ual producer's situation will have to  be developed.   The
                         recommendations  presented here describe general situa-
                         tions.   The number  of replicates and sampling frequency
                         also must be planned prior to sampling.
    
                8.3.2     If possible, mix the source  thoroughly before collecting
                         the sample.   If  mixing  is impractical, the sample should
                         be collected from a representative area of the source.
                         If the  liquid is flowing through an  enclosed  system,  sam-
                         pling through a  valve should be  randomly timed.
    
                8.3.3     Fill the bottle  with water,  add  preservative  (Section
                         8.4), cap tightly,  and  shake well.   To prevent the caps
                         from working loose  during storage tape the caps on with  A
                         a  water-insoluble tape.
    
          8.4   Sample preservation  - Samples should  be stored  at 4°C.  Since
                there  is a  possibility of microbial degradation, addition of H2S04
                during collection to a pH <  2 is recommended.  A test  strip is
                used to monitor  the  pH.   Storage times in excess of 4  weeks are
                not recommended.
    
                If residual chlorine is present  in the sample,  it should be
                quenched with sodium thiosulfate.  EPA Methods  330.4 and 330.5
                may be used to measure the residual chlorine.3  Field  test kits
                are available for this purpose.
    
    
    9.0   Sample Preparation
    
          9.1   Sample homogenization and subsampling - The sample  is  homogenized
                by shaking, blending, or  other appropriate mechanical  technique,
                if necessary. If the density of the  sample is  not  between 0.9
    
    
                                       D-21
    

    -------
          and 1.1, the density should be determined and reported.   Consider-
          ation should be given to treating the sample as a product waste
          (see separate protocol).
    
          Note:    The precision of the mass determination at this step will
                   be reflected in the overall method precision.   Therefore,
                   an analytical balance must be used to assure that the
                   weight is accurate to ±1% or better.
    
    9.2   Surrogate addition - An appropriate volume of surrogate  solution
          SSxxx is pipetted into the sample.  The final concentration of the
          surrogates must be in the working range of the calibration and
          well above the matrix background.
    
          Note:    The volume measurement of the spiking solution  is criti-
                   cal to the overall method precision.   The analyst must
                   exercise caution that the volume is known to ±1% or
                   better.  Where necessary, calibration of the pipet is
                   recommended.
    
    9.3   Sample preparation (extraction/cleanup) - After addition of the
          surrogates, the sample is further treated at the discretion of
          the analyst, provided that the GC/EIMS response of the four sur-
          rogates meets the criteria listed in Section 7.0.   The literature
          pertaining to these techniques has been reviewed.4  Several pos-
          sible techniques are presented below for guidance only.   The ap-
          plicability of any of these techniques to a specific sample matrix
          must be determined by the precision and accuracy of the    C PCB
          surrogate recoveries, as discussed in Section 14.2.
    
          9.3.1    Extraction - The entire sample must be transferred to the
                   extraction vessel with PCB-free water washing,  if neces-
                   sary, to transfer all solids.  The container is then
                   rinsed with the extraction solvent to recovery  any PCBs
                   adhering to' the container wall.  The solvent rinses are
                   combined with the extracts from below.   Measure the sam-
                   ple volume to the nearest 0.5%.
    
                   9.3.1.1  Liquid-liquid extraction - The solvent,  number
                            of extractions, solvent-to-sample ratio, and
                            other parameters are chosen at the analyst's
                            discretion.   A suggested extraction from water
                            is presented in EPA Methods  6081 and 625.2
    
                   9.3.1.2  Sorbent column extraction -  PCBs may be isolated
                            from water onto sorbent columns, although these
                            techniques are not as widely used or thoroughly
                            validated as liquid-liquid extraction.   The
                            selection of sorbent (XAD, Porapak, carbon-
                            polyurethane foam, etc.) will depend on the
                            nature of the matrix.  The available methods
                            have been reviewed.4
                                 D-22
    

    -------
    9.3.2    Cleanup - Several tested cleanup techniques are described
             below.  All but the base cleanup (9.3.2.8)  were previously
             validated for PCBs in transformer fluids.5   Depending
             upon the complexity of the sample,  one or more of the
             techniques may be required to fractionate the PCBs from
             interferences.  For most cleanups a concentrated (1-5
             ml) extract should be used.
    
             9.3.2.1  Acid cleanup
    
                      9.3.2.1.1  Place 5  ml of concentrated sulfuric
                                 acid into a 40-ml narrow-mouth screw-
                                 cap bottle.  Add the sample extract.
                                 Seal the bottle with a  Teflon-lined
                                 screw cap and shake for 1 min.
    
                      9.3.2.1.2  Allow the phases to separate, trans-
                                 fer the  sample  (upper phase) with
                                 three rinses of 1-2 ml  solvent to a
                                 clean container and concentrate to
                                 an appropriate  volume.
    
                      9.3.2.1.3  Analyze  as described in Section 10.0.
    
                      9.3.2.1.4  If the sample is highly contaminated,
                                 a second or third acid  cleanup may
                                 be employed.
    
             9.3.2.2  Florisil column cleanup
    
                      9.3.2.2.1  Variations among batches of Florisil
                                 (PR grade or equivalent) may affect
                                 the elution volume of the various
                                 PCBs. For this reason, the volume
                                 of solvent required to  completely
                                 elute all of the PCBs must be veri-
                                 fied by  the analyst. The weight of
                                 Florisil can then be adjusted accor-
                                 dingly.
    
                      9.3.2.2.2  Place a  20-g charge of  Florisil,
                                 activated overnight at  130°C, into a
                                 Chromaflex column.   Settle the Flor-
                                 isil by  tapping the column.  Add
                                 about 1  cm of anhydrous sodium sul-
                                 fate to  the top of the  Florisil.
                                 Pre-elute the column with 70-80 ml
                                 of hexane.  Just before the exposure
                                 of the sodium sulfate layer to air,
                                 stop the flow.   Discard the eluate.
                           D-23
    

    -------
             9.3.2.2.3  Add the sample extract to the column.
    
             9.3.2.2.4  Carefully wash down the inner wall
                        of the column with 5 ml of the hexane.
    
             9.3.2.2.5  Add 220 ml of hexane to the column.
    
             9.3.2.2.6  Discard the first 25 ml.
    
             9.3.2.2.7  Collect 200 ml of hexane eluate in a
                        Kuderna-Danish flask.   All of the
                        PCBs should be in this fraction.
                        Concentrate to an appropriate volume.
    
             9.3.2.2.8  Analyze the sample as  described in
                        Section 10.0.
    
    9.3.2.3  Alumina column cleanup
    
             9.3.2.3.1  Adjust the activity of the alumina
                        (Fisher A540 or equivalent) by heat-
                        ing to 200°C for 2 to  4 hr.  When
                        cool,  add 3% water (wt:wt) and mix
                        until  uniform.   Store  in a tightly
                        sealed bottle.   Allow  the deactivated
                        alumina to equilibrate at least 1/2
                        hr before use.   Reactivate weekly.
    
             9.3.2.3.2  Variations between batches of alumina
                        may affect the elution volume of  the
                        various PCBs.   For this reason,  the
                        volume of solvent required to com-
                        pletely elute all of the PCBs must
                        be verified by the analyst.  The
                        weight of alumina can  then be ad-
                        justed accordingly.
    
             9.3.2.3.3  Place  a 50-g charge of alumina into
                        a Chromaflex column.   Settle the  alu-
                        mina by tapping.   Add  about 1 cm  of
                        anhydrous sodium sulfate.   Pre-elute
                        the column with 70-80  ml of hexane.
                        Just before exposure of the sodium
                        sulfate layer to air,  stop the flow.
                        Discard the eluate.
    
             9.3.2.3.4  Add the sample extract to the column.
    
             9.3.2.3.5  Carefully wash down the inner wall
                        of the column with 5 ml volume of
                        hexane.
                  D-24
    

    -------
             9.3.2.3.6  Add 295 ml of hexane to the column.
    
             9.3.2.3.7  Discard the first 50 ml.
    
             9.3.2.3.8  Collect 250 ml of the hexane in a
                        Kuderna-Danish flask.  All of the
                        PCBs should be in this fraction.
                        Concentrate to an appropriate volume.
    
             9.3.2.3.9  Analyze the sample as described in
                        Section 10.0.
    
    9.3.2.4  Silica gel column cleanup
    
             9.3.2.4.1  Activate silica gel (Davison grade
                        950 or equivalent) at 135°C overnight.
    
             9.3.2.4.2  Variations between batches of silica
                        gel may affect the elution volume of
                        the various PCBs.   For this reason,
                        the volume of solvent required to
                        completely elute all of the PCBs  must
                        be verified by the analyst.  The
                        weight of silica gel can then be  ad-
                        justed accordingly.
    
             9.3.2.4.3  Place a 25-g charge of activated
                        silica gel into a Chromaflex column.
                        Settle the silica gel by tapping  the
                        column.  Add about 1 cm of anhydrous
                        sodium sulfate to the top of the
                        silica gel.
    
             9.3.2.4.4  Pre-elute the column with 70-80 ml
                        of hexane.  Discard the eluate.  Just
                        before exposing the sodium sulfate
                        layer to air, stop the flow.
    
             9.3.2.4.5  Add the sample extract to the column.
    
             9.3.2.4.6  Wash down the inner wall of the column
                        with 5 ml of hexane.
    
             9.3.2.4.7  Elute the PCBs with 195 ml of 10%
                        diethyl ether in hexane (v:v).
    
             9.3.2.4.8  Collect 200 ml of the eluate in a
                        Kuderna-Danish flask.  All of the
                        PCBs should be in this fraction.
                        Concentrate to an appropriate volume.
                  D-25
    

    -------
             9.3.2.4.9  Analyze the sample according to Sec-
                        tion 10.0.
    
    9.3.2.5  Gel permeation cleanup
    
             9.3.2.5.1  Set up and calibrate the gel perme-
                        ation chromatograph with an SX-3
                        column according to the Autoprep in-
                        struction manual.  Use 15% methylene
                        chloride in cyclohexane (v:v) as the
                        mobile phase.
    
             9.3.2.5.2  Inject 5.0 ml  of the sample extract
                        into the instrument.  Collect the
                        fraction containing the PCBs (see
                        Autoprep operator's manual) in a
                        Kuderna-Danish flask equipped with
                        a 10-ml ampul.
    
             9.3.2.5.3  Concentrate the PCB fraction to an
                        appropriate volume.
    
             9.3.2.5.4  Analyze as described in Section 10.0.
    
    9.3.2.6  Acetonitrile partition
    
             9.3.2.6.1  Place the sample extract into a 125-ml
                        separatory funnel with enough hexane
                        to bring the final volume to 15 ml.
                        Extract the sample four times by shak-
                        ing vigorously for 1 min with 30-ml
                        portions of hexane-saturated acetoni-
                        trile.
    
             9.3.2.6.2  Combine and transfer the acetonitrile
                        phases  to a 1-liter separatory funnel
                        and add 650 ml of distilled water
                        and 40  ml of saturated sodium chloride
                        solution.  Mix thoroughly for about 30
                        sec. Extract  with two 100-ml por-
                        tions of hexane by vigorously shaking
                        about 15 sec.
    
             9.3.2.6.3  Combine the hexane extracts in a
                        1-liter separatory funnel and wash
                        with two 100-ml portions of distilled
                        water.   Discard the water layer and
                        pour the hexane layer through a 8-10
                        cm anhydrous sodium sulfate column
                        into a  500-ml  Kuderna-Danish flask
                        equipped with  a 10-ml ampul.  Rinse
                        the separatory funnel and column with
                        three 10-ml portions of hexane.
    
                  D-26
    

    -------
             9.3.2.6.4  Concentrate the extracts to an
                        appropriate volume.
    
             9.3.2.6.5  Analyze as described in Section 10.0.
    
    9.3.2.7  Florisil slurry cleanup
    
             9.3.2.7.1  Place the sample extract into a 20-ml
                        narrow-mouth screw-cap container.
                        Add 0.25 g of Florisil (PR grade or
                        equivalent).  Seal with a Teflon-lined
                        screw cap and shake for 1 min.
    
             9.3.2.7.2  Allow the Florisil to settle; then
                        decant the treated solution into a
                        second container with rinsing.  Con-
                        centrate the sample to an appropriate
                        volume.  Analyze as described in Sec-
                        tion 10.0.
    
    9.3.2.8  Base cleanup6
    
             9.3.2.8.1  Quantitatively transfer the concen-
                        trated extract to a 125-ml extraction
                        flask with the aid of several small
                        portions of solvent.
    
             9.3.2.8.2  Evaporate the extract just to dry-
                        ness with a gentle stream of dry
                        filtered nitrogen, and add 25 ml of
                        2.5% alcoholic KOH.
    
             9.3-2.8.3  Add a boiling chip,  put a water con-
                        denser in place, and allow the solu-
                        tion to reflux on a hot plate for 45
                        min.
    
             9.8.2.8.4  After cooling, transfer the solution
                        to a 250-ml separatory funnel with
                        25 ml of distilled water.
    
             9.3.2.8.5  Rinse the extraction flask with 25
                        ml of hexane and add it to the
                        separatory funnel.
    
             9.3.2.8.6  Stopper the separatory funnel and
                        shake vigorously for at least 1 min.
                        Allow the layers to separate and
                        transfer the lower aqueous phase to
                        a second separatory funnel.
                  D-27
    

    -------
                                  9.3.2.8.7  Extract the saponification solution
                                             with a second 25-ml portion of hexane.
                                             After the layers have separated,  add
                                             the first hexane extract to the sec-
                                             ond separatory funnel and transfer
                                             the aqueous alcohol layer to the
                                             original separatory funnel.
    
                                  9.3.2.8.8  Repeat the extraction with a third
                                             25-ml portion of hexane.   Discard
                                             the saponification solution, and  com-
                                             bine the hexane extracts.
    
                                  9.3.2.8.9  Concentrate the hexane layer to an
                                             appropriate volume and analyze ac-
                                             cording to Section 10.0.
    
    
    10.0  Gas Chromatographic/Electron Impact Mass Spectrometric Determination
    
          10.1  Internal standard addition -  An appropriate volume of  the internal
                standard solution is  pipetted into the sample.   The final concen-
                tration of the internal standard must be in the working range  of
                the calibration and well above the matrix background.   The inter-
                nal standard is thoroughly incorporated by mechanical  agitation.
    
                Note:   The volumetric measurement of the internal standard solu-
                tion is critical to the overall method precision.   The  analyst
                must exercise caution that the volume is known to be ±1% or better.
                Where necessary, calibration  of the pipet is recommended.
    
          10.2  Tables 2,  and 5 through 8 summarize the recommended operating  con-
                ditions for analysis.   Figure 1 presents an example of  a chromato-
                gram.
    
          10.3  While the  highest available chromatographic resolution  is not  a
                necessary  objective of this protocol,  good chromatographic per-
                formance is recommended.   With the high resolution of  CGC, the
                probability that the  chromatographic peaks consist of  single com-
                pounds is  higher than with PGC.   Thus,  qualitative and  quantita-
                tive data  reduction should be more reliable.
    
          10.4  After performance of  the system has been certified for  the day
                and all instrument conditions set according to Tables  2, and 5
                through 8,  inject an  aliquot  of the sample onto the GC  column.
                If the response for any ion,  including surrogates and  internal
                standards,  exceeds the working range of the system,  dilute the
                sample and reanalyze.   If the responses of surrogates,  analyte,
                or internal standard  are below the working range,  recheck the
                system performance.   If necessary,  concentrate the sample and
                reanalyze.
                                       D-28
    

    -------
             108.On
               RIG
    (V)
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    rH
    
    C
    0)
    x;
    p.
    •H
    XI
    o
    )_l
    o
    rH
    X!
    o
    o
    G
    o
    a
    1
    CN
    
    G10
    1
    523 575 1 1
    
    033
    10:0;)
    vO
    CJ
    CO
    rH
    rH rH
    C C
    01 01
    X! XI
    P. P.
    iH -H
    XI XI
    0 0
    r4 t-l
    O O
    rH rH
    x; x:
    o o
    o o
    c c
    0 0
    
    1 1
    
    •o
    4-J l^""K
    CO 01
    4-1
    rH CO
    cd oo
    c o
    01 M
    4-1 3
    C to
    
    vO CO
    •O rH
    1 1
    rH rH
    t^t r^»
    C C
    Ol Ol
    XI XI
    ex p.
    •H -iH
    Xi Xi
    o o
    r^lfVV^
    o
    1 — 1
    XI
    o
    cd
    J^
    4-1
    Ol
    H
    1
    
    rH -4T
    rS •*
    c -*
    0) -
    x: -
    (X CO
    •iH
    Xi co
    0
    
    o
    rH
    XI
    CJ
    cd
    4-1
    c
    01
    i
    •>
    "^
    ro
    CN" I
    -J
    
    rH
    XI
    O
    Cd rH
    (-1 >i
    4J C
    01 01
    H XI
    1 CX
    •H
    ST XI
    « O
    -3- l-l
    - O
    - rH
    ro XI
    • O
    en cd
    X
    01
    32
    I
    —
    vD
    "
    "1
    -*
    CO
    CN
    CN
    
    1101
    L
    iiwe
    IGHU
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    rH
    
    B
    O)
    XI
    cx
    •H
    
    O
    
    o
    rH
    XI
    O
    cd
    4-1
    p.
    01
    33
    
    vO
    V.
    •.
    LO
    •»
    -
    ^
    ~*
    n
    CM K
    CN"
    •
    1178
    1
    20-
    
    
    
    
    
    
    
    
    
    m
    
    rH
    ^
    rH C
    >-, 01
    C XI
    0) CX
    X! -H rH
    P. XI >.
    •H O C
    XI MO)
    o ox;
    t-l rH P.
    o x: -H
    rH O XI
    xi cd o
    U C t-i
    cd o o
    4-1 a rH
    u i x:
    O 0
    I vo cd
    o
    vO vo Ol
    n f. Q
    VO 1A
    •t n
    f ~
    if) -^
    * •*
    ir> ^f
    «t n
    •. •.
    ro ro
    A *
    ro ro
    
    •• •.
    CN CN
    CN CN
    
    13IG
    
    
    
    JL
    ' . 112592.
    
    CN
    rH
    0
    ro •
    rH
    |
    rH
    ^
    C
    Ol
    x:
    P-. *-\
    •H Ol
    n ^j
    O cd
    t-i 60
    0 O
    *~H VH
    _<^ ^
    0 3
    cd 10
    O — '
    Ol
    Q
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    •
    0 l'i'i:9 SCA'l
    Ct) 23:20 Tllh:
         Figure  1.  Capillary gas chromatography/electron  impact  ionization  mass  spectrometry (CGC/EIMS)
           chromatogram or  the calibration  standard  solution  required  for  quantitation  of  PCBs by homolog.
           This  chromatogram includes PCBs  representative  of  each homolog,  three  carbon-13 labeled surrogates,
           and the deuterated internal  standard.  The  concentration  of all components and  the CGC/EIMS
           parameters are presented  in  Tables  3, 4,  5, and 7.
    

    -------
          10.5  Record all data on a digital storage device (magnetic disk,  tape,
                etc.)  for qualitative and quantitative data reduction as discussed
                Vx /-s 1 ^\r.ir
                below.
    11.0  Qualitative Identification
    
          11.1  Selected ion monitoring (SIM)  or limited mass scan (IMS)  data -
                The identification of a compound as a given PCB homolog requires
                that two criteria be met:
    
                11.1.1   (1) The peak must elute within the retention time window
                         set for that homolog  (Section 7.5);  and (2)  the  ratio of
                         two ions obtained by  SIM (Table 11)  or by IMS (Table 12)
                         must match the natural ratio within ±20%.   The analyst
                         must search the higher mass windows, in particular M+70,
                         to prevent misidentification of a PCB fragment ion clus-
                         ter as the parent.
    
                11.1.2   If one or the other of these criteria is not met, inter-
                         ferences may have affected the results and a reanalysis
                         using full scan EIMS  conditions is recommended.
    
          11.2  Full scan data
    
                11.2.1   The peak must elute within the retention time windows
                         set for that homolog  (as described in Section 7.5).
    
                11.2.2   The unknown spectrum  must match that of an authentic PCB.
                         The intensity of the  three largest ions in the molecular
                         cluster (two largest  for monochlorobiphenyls) must match
                         the natural ratio within ±20%.  Fragment clusters with
                         proper intensity ratios must also be present.
    
                11.2.3   Alternatively, a spectral search may be used to  automat-
                         ically reduce the data.  The criteria for acceptable
                         identification include a high index of similarity.  For
                         the Incos 2300, a fit of 750 or greater must be  obtained.
    
          11.3  Disputes in interpretation - Where there is reasonable doubt as
                to the identity of a peak as a PCB, the analyst must either iden-
                tify the peak as a PCB or proceed to a confirmational analysis
                (see Section 13.0).
    12.0  Quantitative Data Reduction
    
          12.1  Once a chromatographic peak has been identified as a PCB, the com-
                pound is quantitated based either on the integrated abundance of
                the SIM data or EICP for the primary characteristic ion in Tables
                11 and 12.   If interferences are observed for the primary ion,
                                       D-30
    

    -------
                     TABLE 11.  CHARACTERISTIC SIM IONS FOR PCBs
    
    Homolog
    Ci^HgCl
    Cl2HsCl2
    C^HyCla
    C12H6C14
    C12H5Cl5
    ^12^4^16
    C12H3Cl7
    C12H2C18
    Ci2HCl9
    CiaCliO
    
    Primary
    188 (100)
    222 (100)
    256 (100)
    292 (100)
    326 (100)
    360 (100)
    394 (100)
    430 (100)
    464 (100)
    498 (100)
    Ion (relative intensity)
    Secondary
    190 (33)
    224 (66)
    258 (99)
    290 (76)
    328 (66)
    362 (82)
    396 (98)
    432 (66)
    466 (76)
    500 (87)
    
    Tertiary
    -
    226 (11)
    260 (33)
    294 (49)
    324 (61)
    364 (36)
    398 (54)
    428 (87)
    462. (76)
    496 (68)
    
    Source:  Rote, J.  W.,  and W.  J.  Morris,  "Use of Isotopic Abundance Ratios in
             Identification of Polychlorinated Biphenyls by Mass Spectroraetry,"
             J. Assoc. Offic. Anal.  Chem.,  56(1), 188-199 (1973).
                                         D-31
    

    -------
                TABLE 12.  LIMITED MASS SCANNING (IMS) RANGES FOR PCBs
    
    Compound
    C^HgCli
    C12H8C12
    C12H7C13
    C12H6C13
    C^Hsds
    C12H4Cl6
    C12H3Cl7
    C12H2C18
    Cl2HClg
    Cl2cll6
    C12D6C14
    ^Cg^CgHgCl
    13C12H6C14
    13C12H2C18
    "C12C110
    Mass range (m/z)
    186-190
    220-226
    254-260
    288-294
    322-328
    356-364
    386-400
    426-434
    460-468
    494-504
    294-300
    192-1.96
    300-306
    438-446
    506-516
    
    a  Adapted from Tindall, G.  W.,  and P.  E.  Wininger,  "Gas Chromatography-Mass
       Spectrometry Method for Identifying and Determining Polychlorinated Bi-
       phenyls," J. Chromatogr., 196,  109-119  (1980).
                                         D-32
    

    -------
          use the secondary and then tertiary ion for quantitation.   If in-
          terferences  in the parent cluster prevent quantitation,  an ion
          from a fragment cluster (e.g.,  M-70) may be used.   Whichever ion
          is used,  the RF must be determined using that ion.   The  same cri-
          teria should be applied to the  surrogate compounds  (Table  13).
    
    12.2  Using the appropriate analyte-internal standard pair and response
          factor (RF ) as determined in Section 7.3, calculate the concen-
          tration of*each peak using Equation 12-1.
    
                                      A           Mig   Vg
               Concentration (pg/g) = •£- ' gjr- • ^ • ^        Eq.  12-1
                                       is     p    e     i
    
          where     A  = area of the characteristic ion for  the analyte PCB
                     p     peak
    
                   A.   = area of the characteristic ion for  the internal
                           standard peak
    
                   RF  = response factor  of a given PCB congener
    
                   M.   = mass of internal standard injected  (micrograms)
                    IS
                    M  = mass of sample extracted (grams)
    
                    V. = volume injected  (microliters)
    
                    V  = volume of sample extract (microliters)
    
    12.3  If a peak appears to contain non-PCB interferences  which cannot
          be circumvented by a secondary  or tertiary ion, either:
    
          12.3.1   Reanalyze the sample on a different column which  sepa-
                   rates the PCB and interferents;
    
          12.3.2   Perform additional chemical cleanup (Section 9) and then
                   reanalyze the sample;  or
    
          12.3.3   Quantitate the entire  peak as PCB.
    
    12.4  Calculate the recovery of the four 13C surrogates  using  the  ap-
          propriate surrogate-internal standard pair and response  factor
          (RF. ) as determined in Section 7.4 using Equation 12-2.
             1S
                                      A           M.
                       Recovery (%) = -f- • ^- - —^ - 100        Eq.  12-2
                                       is     s    s
          where A  = area of the characteristic ion for the  surrogate  peak
                 S
               A.  = area of the characteristic ion for the  internal standard
                is        ,
                       peak
                                 D-33
    

    -------
            TABLE 13.  CHARACTERISTIC IONS FOR 13C-LABELED PCB SURROGATES
    
    
    
                                  	Ion (relative intensity)	
    Specific compound              Primary           Secondary          Tertiary
    
    
    
    
    
    
    i3C612C6H9Cl                  194 (100)          196 (33)
    
    
    
    13C12H6C14                    304 (100)          306 (49)           302  (78)
    
    
    
    13C12H2C18                    442 (100)          444 (65)           440  (89)
    
    
    
    13C12C110                     510 (100)          512 (87)           514  (50)
                                         D-34
    

    -------
                     RF  = response factor for the surrogate compound with respect
                       s     to the internal  standard (Equation 7-2)
    
                     M.   = mass of internal standard injected (nanograms)
                      IS
    
                      M  = mass of surrogate,  assuming 100% recovery  (nanograms)
                       S
          12.5  Correct the concentration of  each peak using Equation 12-3.   This
                is the final reportable concentration.
    
    
          Corrected concentration (Mg/g)  = C°nRecovery°(%^/S * 10°     Eq*  12'3
    
          12.6  Sum all of the peaks for  each homolog,  and then sum those  to yield
                the total PCB concentration in the sample.  Report all numbers in
                [Jg/g.  The reporting form in  Table 14 may be used.  If an  alter-
                nate reporting format (e.g.,  concentration per peak)  is desired,
                a different report form may be used.  The uncorrected concentra-
                tions, percent recovery,  and  corrected recovery are to be  reported.
    
          12.7  Round off all numbers reported to two significant figures.
    
    
    13.0  Confirmation
    
          If there is reason to question the  qualitative identification (Section
          11.0), the analyst may choose to confirm that a peak is not a PCB.  Any
          technique may be chosen provided that it is validated as having  equiva-
          lent or superior selectivity and sensitivity to GC/EIMS.  Some candidate
          techniques include alternate GC columns (with EIMS detection), GC/CIMS,
          GC/NCIMS, high resolution EIMS, and MS/MS techniques.  Each laboratory
          must validate confirmation techniques to show equivalent or superior
          selectivity between PCBs and interferences and sensitivity  (limit of
          quantitation, LOQ).
    
          If a peak is confirmed as being a non-PCB, it may be deleted from the
          calculation (Section 12).  If a peak is confirmed as containing  both
          PCB and non-PCB components, it must be quantitated according to  Section
          12.3.
    14.0  Quality Control
    
          14.1  Each laboratory that uses this method must operate a formal qual-
                ity control (QC) program.  The minimum requirements of this pro-
                gram consist of an initial demonstration of laboratory capability
                and the analysis of spiked samples as a continuing check on per-
                formance.  The laboratory must maintain performance records to
                define the quality of data that are generated.   After a date spe-
                cified by the Agency, ongoing performance checks should be com-
                pared with established performance criteria to  determine if the
                results of analyses are within accuracy and precision limits ex-
                pected of the method.
    
                                       D-35
    

    -------
                                 TABLE 14.  ANALYSIS REPORT
                                INCIDENTAL PCBs IN WASTEWATER
    Sample No. 	
    Sample Matrix 	
    Sample Source 	
    Notebook No. or File Location
    Volume Extracted 	 liter
    Extraction/Cleanup Procedure	
    Int. Std.       Mass Added (|Jg)      (Circle one)       Ratio       Intensity
    
    4-Cl(d6)                              298    300      100/49
    
    
    
    Surrogates   Mass Added (|jg)    (Circle one)     Ratio     Intensity   % Recovery
    
      1-C1                           194    196     100/33
    
      4-C1                           304    306     100/49
    
      8-C1                           442    444     100/65
    
     10-C1                           510    512     100/87
                                         (continued)
                                           D-36
    

    -------
    TABLE 14 (continued)
    
    Qualitative Quantitative
    Analyte 1° 2° Il°
    1-C1 188 190
    2-C1 222 224
    3-C1 256 258
    4-C1 292 290
    5-C1 326 328
    6-C1 360 362
    7-C1 394 396
    8-C1 430 432
    9-C1 464 466
    10-C1 498 500
    Total
    Reported by:
    Name
    Signature/Date
    Organization
    Uncorr Corr
    Ion Cone. Cone
    12° Ratio Theoretical OK? Used RF (Mg/£) (Mg/£)
    100/33
    100/66
    100/99
    100/76
    100/66
    100/82
    100/98
    100/66
    100/76
    100/87
    MS/£ M8/£
    Uncorr. Corr.
    Internal Audit: EPA Audit:
    Name Name
    Signature/Date Signature/Date
    Organization Organization
         D-37
    

    -------
    14.2  The analysts must certify that the precision and accuracy of the
          analytical results are acceptable by:
    
          14.2.1   The absolute precision of surrogate recovery, measured
                   as the RSD  of the integrated EIMS area (A ) for a set
                   of samples, must be ±10%.
    
          14.2.2   The mean recovery (R ) of at  least four replicates of a
                   QC check sample to be supplied by the Agency must meet
                   Agency-specified accuracy and precision criteria.  This
                   forms the initial data base for establishing control
                   limits (see Section 14.3 below).
    
    14.3  Control limits - The laboratory must establish control limits
          using the following equations:
    
                   Upper control limit (UCL) = RC + 3 RSDc
    
                   Upper warning limit (UWL) = R  +2 RSD
    
                   Lower warning limit (LWL) = R  - 2 RSD
    
                   Lower control limit (LCL) = R  - 3 RSD
    
          These may be plotted on control charts.   If an analysis of a
          check sample falls outside the warning limits, the analyst should
          be alerted that potential problems may need correction.  If the
          results for a check sample fall outside the control limits, the
          laboratory must take corrective action and recertify the perfor-
          mance (Section 14.2) before proceeding with analyses.   The warn-
          ing and control limits should be continuously updated as more
          check sample replicates are added to the data base.
    
    14.4  Before processing any samples, the analyst should demonstrate
          through the analysis of a reagent blank that all glassware and
          reagent interferences are under control.   Each time a set of sam-
          ples is analyzed or there is a change  in reagents,  a laboratory
          reagent blank should be processed as a safeguard against con-
          tamination.
    
    14.5  Procedural QC - The various steps of the analytical procedure
          should have quality control measures.   These include but are not
          limited to:
    
          14.5.1   GC performance - See Section  7.1  for performance cri-
                   teria.
    
          14.5.2   MS performance - See Section  7.2  for performance cri-
                   teria.
                                 D-38
    

    -------
                14.5.3    Qualitative  identification - At  least  10% of the PCB
                         identifications,  as well  as any  questionable results,
                         should be  confirmed by  a  second  mass spectrometrist.
    
                14.5.4    Quantitation - At least 10% of all manual calculations,
                         including  peak area calculations, must be checked.  After
                         changes  in computer quantitation routines,  the  results
                         should be  manually checked.
    
          14.6  A minimum of  10%  of all samples, one sample per month or one sam-
                ple  per  matrix type,  whichever is  greater, selected  at random,
                must be  run in triplicate  to monitor the  precision of the analy-
                sis.  An RSD  of ±30%  or less must  be achieved.  If the precision
                is greater than ±30%,  the  analyst  must be recertified (see Section
                14.2).
    
          14.7  A minimum of  10%  of all samples, one sample per month or one sam-
                ple  per  matrix type,  whichever is  greater, selected  at random,
                must be  analyzed  by the standard addition technique.  Two aliquots
                of the sample are analyzed, one  "as is" and one spiked (surrogate
                spiking  and equilibration  techniques are  described in Section 9.2)
                with a sufficient amount of Solution CSxxx to yield  approximately
                100  (jg/liter  of each  compound).  The samples are  analyzed together
                and  the  quantitative  results calculated.  The recovery of the
                spiked compounds  (calculated by  difference) must  be  80-120%.  If
                the  sample is known to contain specific PCB isomers, these isomers
                may  be substituted  for solution  CSxxx.  If the  concentrations of
                PCBs are known to be  high  or low,  the amount added should be ad-
                justed so that the  spiking level is 1.5 to 4 times the measured
                PCB  level in  the  unspiked  sample.
    
          14.8  Interlaboratory comparison - Interlaboratory comparison  studies
                are  planned.  Participation requirements, level of performance,
                and  the  identity  of the coordinating laboratory will be  presented
                in later revisions.
    
          14.9  It is  recommended that the participating  laboratory  adopt addi-
                tional QC practices for use with this method.   The specific prac-
                tices  that are most productive depend upon the  needs of  the lab-
                oratory  and the nature of  the samples.  Field duplicates or
                triplicates may be  analyzed to monitor the precision of  the sam-
                pling  technique.  Whenever possible, the  laboratory  should per-
                form analysis of  standard  reference materials and participate in
                relevant performance  evaluation  studies.
    
    
    15.0  Quality Assurance
    
          Each participating  laboratory must develop a quality  assurance plan ac-
          cording to EPA guidelines.7 The quality assurance plan must be submitted
          to the Agency  for approval.
                                       D-39
    

    -------
    16.0  Method Performance
    
          The method performance is being evaluated.   Limits of quantitation;
          average intralaboratory recoveries,  precision,  and accuracy;  and inter-
          laboratory recoveries, precision,  and accuracy  will be presented.
    
    
    17.0  Documentation and Records
    
          Each laboratory is responsible for maintaining  full records of the analy-
          sis.  Laboratory notebooks should  be used for handwritten records.  GC/MS
          data must be archived on magnetic  tape,  disk, or a similar device.  Hard
          copy printouts may be kept in addition if desired.  QC records should
          be maintained separately from sample analysis records.
    
          The documentation must describe completely how  the analysis was performed.
          Any variances from the protocol must be  noted and fully described.  Where
          the protocol lists options (e.g.,  sample cleanup), the option used and
          specifics (solvent volumes, digestion times, etc.) must be stated.
                                       D-40
    

    -------
                                     REFERENCES
    
    1.  Environmental Protection Agency,  Organochlorine Pesticides and PCBs—
        Method 608," Fed.  Reg., 44,  69501-69509 (December 3,  1979).
    
    2.  Environmental Protection Agency,  "Base/Neutrals, Acids,  and Pesticides--
        Method 625," Fed.  Reg., 44,  69540-69552 (December 3,  1979),  and subse-
        quent revisions.
    
    3.  "Methods 330.4 (Titrimetric, DPD-FAS)  and 330.5 (Spectrophotometric,  DPD)
        for Chlorine, Total Residual," Methods for Chemical Analysis of Water and
        Wastes, U.S. Environmental Protection Agency,  Environmental Monitoring and
        Support Laboratory, Cincinnati, Ohio,  March 1979, EPA 600-4/79-020.
    
    4.  Erickson, M. D.,  and J. S. Stanley,  "Methods of Analysis for Incidentally
        Generated PCBs Literature Review and Preliminary Recommendations," Interim
        Report No. 1, EPA Contract No. 68-01-5915, Task 51, 1982.
    
    5.  Bellar, T. A., and J. J. Lichtenberg,  "The Determination of Polychlorinated
        Biphenyls in Transformer Fluid and Waste Oils," Prepared for U.S.  Environ-
        mental Protection Agency (1981).   EPA-600/4-81-045.
    
    6.  American Society for Testing and Materials, "Standard Method for Analysis
        of Environmental Materials for Polychlorinated Biphenyls," pp.  877-885,
        in Annual Book of ASTM Standards, Part 40, Philadelphia, Pennsylvania
        (1980).  ANSI/ASTM D 3304-77.
    
    7.  Quality Assurance Program Plan for the Office  of Toxic Substances, Office
        of Pesticides and Toxic Substances,  U.S. Environmental Protection Agency,
        Washington, D.C.,  October 1980.
                                       D-41
    

    -------
                                       TECHNICAL REPORT DATA
                                (Please read Instructions on the reverse before completing)
    1. REPORT NO.
    EPA-560/5-82-006
                                 2.
                                                               3. RECIPIENT'S ACCESSION NO.
    4. TITLE AND SUBTITLE
    Analytical Methods for By-Product PCBs—Initial
       Validation and Interim Protocols
                                  5. REPORT DATE
                                    October 11, 1982
                                  6. PERFORMING ORGANIZATION CODE
    7. AUTHOR(S)  Mitchell D. Erickson, John  S.  Stanley,  Gil
    Radolovich,  Kay Turman, Karin Bauer,  Jon Onstot,  Donna
    Rose,  and  Margaret Wickham  	
                                  8. PERFORMING ORGANIZATION REPORT NO.
                                    MRI Project No. 4901-A51
    9. PERFORMING ORGANIZATION NAME AND ADDRESS
    Midwest Research Institute
    425  Volker Boulevard
    Kansas City,  MO  64110
                                                               10. PROGRAM ELEMENT NO.
                                   11. CONTRACT/GRANT NO.
                                    EPA 68-01-5915, Task  51
    12. SPONSORING AGENCY NAME AND ADDRESS
    U.S.  Environmental Protection Agency
    Office of Toxic Substances, Field  Studies  Branch
    TS-798
    Washington,  DC  20460           	
                                   13. TYPE OF REPORT AND PERIOD COVERED
                                    Interim 4, 4/24-8/31/82   .
                                   14. SPONSORING AGENCY CODE
    15. SUPPLEMENTARY NOTES
    
    The  task manager is David P. Redford;  the  project officer is Frederick W. Kutz.
    16. ABSTRACT
    This  document presents proposed analytical methods for analysis of by-product  PCBs in
    commercial products, product waste  streams,  wastewaters, and air.  The analytical
    method  for commercial products and  product waste streams consist of a flexible approach
    for extraction and cleanup of particular matrices.  The 13C-labeled PCB  surrogates are
    added as part of a strong quality assurance program to determine levels  of  recovery.
    The wastewater method is based on EPA Methods 608 and 625 with revisions to include use
    of the  13c-labeled PCB surrogates.  The air method is a revision of a proposed EPA
    method  for the collection and analysis of  PCBs in air and flue gas emissions.   Capil-
    lary  or packed column gas chromatography/electron impact ionization mass spectrometry
    is proposed as the primary instrumental method.   Response factors and retention times
    of 77 PCB congeners relative to tetrachlorobiphenyl-d6 are presented in  addition to
    statistical analysis to project validity of the data and extrapolation of relative
    response factors to all 209 possible congeners.   Preliminary studies using  the ISC-
    labeled surrogates to validate specific cleanup procedures and to analyze several com-
    mercial products and product wastes indicate that the proposed analytical methods are
    both  feasible and practical.
    17.
                                   KEY WORDS AND DOCUMENT ANALYSIS
                      DESCRIPTORS
                                                 b.IDENTIFIERS/OPEN ENDED TERMS
                                                c. COSATI field/Group
     Polychlorinated biphenyls
     PCBs
     Incidentally generated
     Analytical protocols
     Air
     Wastewater
     Commercial products	
    Commercial, waste
    Capillary column
    Electron impact
    EIMS
    Response factors
    Relative retention
    Surrogates	
    streams
    gas chromatography
    onization mass spectromet
    
     relative response  factor
      times
    18. DISTRIBUTION STATEMENT
     Unlimited
                                                  19. SECURITY CLASS (ThisReport)
                                                  Unclassified
                                                21. NO. OF PAGES
                                                   243
                     20. SECURITY CLASS (This page)
                     Unclassified
                                                                            22. PRICE
    EPA Form 2220-1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE
    

    -------